diff --git a/documentation/5/.buildinfo b/documentation/5/.buildinfo
index edf215e16..1b6340771 100644
--- a/documentation/5/.buildinfo
+++ b/documentation/5/.buildinfo
@@ -1,4 +1,4 @@
 # Sphinx build info version 1
 # This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done.
-config: 8fcd4cc66e68152844a59c6365600984
+config: e7f4ce9044f550b7edaf1bcaedd3edcc
 tags: 645f666f9bcd5a90fca523b33c5a78b7
diff --git a/documentation/5/_sources/building_networks.rst.txt b/documentation/5/_sources/building_networks.rst.txt
index cdcd022d2..be1710964 100644
--- a/documentation/5/_sources/building_networks.rst.txt
+++ b/documentation/5/_sources/building_networks.rst.txt
@@ -7,7 +7,7 @@ Building networks
 The model
 ---------
 A network model is defined as follows:
-A :class:`.GeNNModel` must be created with a name and a default precision (see \ref floatPrecision)}:
+A :class:`.GeNNModel` must be created with a default precision (see \ref floatPrecision) and a name:
 
 ..  code-block:: python
 
@@ -22,7 +22,7 @@ For example, the single-threaded CPU backend could be manually selected with:
     model = GeNNModel("float", "YourModelName", 
                       backend="single_threaded_cpu")
 
-When running models on a GPU, smaller models may not fully occupy the device. In some scenerios such as gradient-based training and parameter sweeping,
+When running models on a GPU, smaller models may not fully occupy the device. In some scenarios such as gradient-based training and parameter sweeping,
 this can be overcome by runing multiple copies of the same model at the same time (batching in Machine Learning speak).
 Batching can be enabled on a GeNN model with:
 
@@ -35,12 +35,19 @@ Whether state variables are duplicated or shared is controlled by the :class:`.V
 associated with each variable. Please see TODO for more details.
 
 Additionally, any preferences exposed by the backend can be configured here. 
-For example, the CUDA backend allows you to select which CUDA device to use via the manual_device_id
+For example, the CUDA backend allows you to select which CUDA device to use via the manual_device_id:
+
+.. code-block:: python
+
+   model = GeNNModel("float", "YourModelName",
+                    backend="cuda", manual_device_id=0)
 
 -----------
 Populations
 -----------
 
+Populations formalise the concept of groups of neurons or synapses that are functionally related or a practical grouping, e.g. a brain region in a neuroscience model or a layer in a machine learning context.
+
 Parameters
 ----------
 Parameters are initialised to constant numeric values which are homogeneous across an entire population:
@@ -49,9 +56,10 @@ Parameters are initialised to constant numeric values which are homogeneous acro
 
     ini = {"m": 0.0529324, ...}
 
-They are very efficient to access from models as their values are either hard-coded into the kernels 
+They are very efficient to access from models as their values are either hard-coded into the backend code 
 or, on the GPU, delivered via high-performance constant cache.
-However 
+However, they can only be used if literally no changes are needed for the entire simulation and all members of
+the population have the exact same parameter value.
 
 Extra global parameters
 -----------------------
@@ -63,25 +71,26 @@ TODO
 
 Variables
 ----------
-Variables can be initialised in many ways.
-By using the GPU to fill them with a constant value:
+Variables contain values that are individual to the members of a population and can change over time. They can be initialised in many ways. The initialisation is configured through a Python dictionary that is then passed to :meth:`.GeNNModel.add_neuron_population` or :meth:`.GeNNModel.add_synapse_population` which create the populations.
+
+To initialise variables one can use the backend, e.g. GPU, to fill them with a constant value:
 
 ..  code-block:: python
 
     ini = {"m": 0.0529324, ...}
 
-by copying in a sequence of values from Python:
+or copy a sequence of values from Python:
 
 ..  code-block:: python
 
     ini = {"m": np.arange(400.0), ...}
 
-or by using a variable initialisation snippet configured using the following function:
+or use a variable initialisation snippet returned by the following function:
 
 .. autofunction:: pygenn.init_var
     :noindex:
 
-and then used in the dictionary the same way:
+The resulting initialisation snippet can then be used in the dictionary in the usual way:
 
 ..  code-block:: python
 
@@ -89,7 +98,7 @@ and then used in the dictionary the same way:
 
 Variables references
 --------------------
-As well as variables and parameters, various types of model have variable references which are used to reference variables belonging to other populations.
+As well as variables and parameters, various types of models have variable references which are used to reference variables belonging to other populations.
 For example, postsynaptic update models can reference variables in the postsynaptic neuron model and custom updates are 'attached' to other populations based on their variable references.
 
 A variable reference called R could be assigned to various types of variable using the following syntax:
@@ -213,4 +222,4 @@ updating model *variables*, they update model *connectivity* (current only with
 Custom connectivity updates are added to a model using:
 
 .. automethod:: .GeNNModel.add_custom_connectivity_update
-    :noindex:
\ No newline at end of file
+    :noindex:
diff --git a/documentation/5/_sources/index.rst.txt b/documentation/5/_sources/index.rst.txt
index cfa5f74c7..ff9e0f103 100644
--- a/documentation/5/_sources/index.rst.txt
+++ b/documentation/5/_sources/index.rst.txt
@@ -12,13 +12,14 @@ Note, this documentation is under construction. If you cannot find what you are
 .. toctree::
     :maxdepth: 4
     :titlesonly:
-    
+
+    installation
     upgrading
     building_networks
     simulating_networks
     custom_models
     bibliography
-    
+
     userproject/index
 
     Reference documentation <source/pygenn>
diff --git a/documentation/5/_sources/installation.rst.txt b/documentation/5/_sources/installation.rst.txt
new file mode 100644
index 000000000..c1a7c6a76
--- /dev/null
+++ b/documentation/5/_sources/installation.rst.txt
@@ -0,0 +1,60 @@
+.. py:currentmodule:: pygenn
+
+============
+Installation
+============
+In future we plan on providing binary builds of GeNN via conda. However, for now, GeNN
+needs to be installed from source.
+
+----------------
+Pre-installation
+----------------
+1.  Install the C++ compiler on the machine, if not already present.
+    For Windows, Visual Studio 2019 or above is required. The Microsoft Visual Studio 
+    Community Edition can be downloaded from
+    https://www.visualstudio.com/en-us/downloads/download-visual-studio-vs.aspx.
+    When installing Visual Studio, one should select the 'Desktop 
+    development with C++' configuration. On Linux, the GNU Compiler 
+    Collection (GCC) 7.5 or above is required. This can be obtained from your
+    Linux distribution repository, for example on Ubuntu by running ``sudo apt-get install g++``, 
+    or alternatively from https://gcc.gnu.org/index.html.
+2.  If your machine has an NVIDIA GPU and you haven't installed CUDA already, 
+    obtain a fresh installation of the NVIDIA CUDA toolkit from
+    https://developer.nvidia.com/cuda-downloads
+    Again, be sure to pick CUDA and C++ compiler versions which are compatible
+    with each other. The latest C++ compiler need not necessarily be
+    compatible with the latest CUDA toolkit.
+3.  GeNN uses the ``CUDA_PATH`` environment variable to determine which 
+    version of CUDA to build against. On Windows, this is set automatically when 
+    installing CUDA. However, if you choose, you can verify which version is 
+    selected by running ``echo %CUDA_PATH`` in a command prompt.
+    However, on Linux and Mac you need to set ``CUDA_PATH`` manually with:
+    ```bash
+    export CUDA_PATH=/usr/local/cuda
+    ```
+    assuming CUDA is installed in /usr/local/cuda (the standard location 
+    on Ubuntu Linux). Again, to make this change persistent, this can
+    be added to your login script (e.g. ``.profile`` or ``.bashrc``)
+4.  Either download the latest release of GeNN and extract into your 
+    home directory or clone using git from https://github.com/genn-team/genn
+5.  On Linux, install the development version of libffi. For example, on Ubuntu you can do this
+    by running ``sudo apt-get install libffi-dev``.
+6.  Install the pybind11, psutil and numpy packages with pip i.e. ``pip install pybind11 psutil numpy``.
+
+----------------------
+Building with setup.py
+----------------------
+From the GeNN directory, the GeNN libraries and python package can be built
+with ``python setup.py install``. If you wish to create an editable install
+(most useful if you are intending to modify GeNN yourself) you can also used
+``python setup.py develop``. On Linux (or Windows if you have a debug version
+of the python libraries installed) you can build a debug version of GeNN with
+``python setup.py build_ext --debug develop``.
+ 
+-----------------
+Building with pip
+-----------------
+From the GeNN directory, the GeNN libraries and python package can be built
+with ``pip install .``. If you wish to create an editable install
+(most useful if you are intending to modify GeNN yourself) you can also used
+``pip install --editable .``.
\ No newline at end of file
diff --git a/documentation/5/_sources/source/pygenn.rst.txt b/documentation/5/_sources/source/pygenn.rst.txt
index a1937c5ce..07419a24e 100644
--- a/documentation/5/_sources/source/pygenn.rst.txt
+++ b/documentation/5/_sources/source/pygenn.rst.txt
@@ -41,6 +41,14 @@ pygenn.custom\_update\_models module
    :undoc-members:
    :show-inheritance:
 
+pygenn.genn module
+------------------
+
+.. automodule:: pygenn.genn
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
 pygenn.genn\_groups module
 --------------------------
 
diff --git a/documentation/5/_sources/upgrading.rst.txt b/documentation/5/_sources/upgrading.rst.txt
index 1db5460c4..0fb1bcf2c 100644
--- a/documentation/5/_sources/upgrading.rst.txt
+++ b/documentation/5/_sources/upgrading.rst.txt
@@ -21,7 +21,11 @@ Syntax changes
 In order to streamline the modelling process and reduce the number of ways to achieve the same thing we have to maintain,
 several areas of GeNN syntax have changed:
 
-- single PSM code string
+- Postsynaptic models were previously implemented with two code strings: ``apply_input_code`` and
+  ``decay_code``. This was unnecessarily complex and the order these were evaluated in wasn't obvious.
+  Therefore, these have been replaced by a single ``sim_code`` code string from which input can be
+  delivered to the postsynaptic neuron using the ``injectCurrent(x)`` function.
+  
 - :meth:`.GeNNModel.add_synapse_population` was becoming rather cumbersome and this was only made worse by postsynaptic models and weight update models now taking variable references.
   To improve this:
 
@@ -30,5 +34,10 @@ several areas of GeNN syntax have changed:
 
 - The row build and diagonal build state variables in sparse/toeplitz connectivity building code were really ugly and confusing. Sparse connectivity init snippets now just let the user write whatever sort of loop they want and do the initialisation outside and toeplitz reuses the for_each_synapse structure described above to do similar.
 - ``GLOBALG`` and ``INDIVIDUALG`` confused almost all new users and were really only used with ``StaticPulse`` weight update models. Same functionality can be achieved with a ``StaticPulseConstantWeight`` version with the weight as a parameter. Then I've renamed all the 'obvious' :class:`.SynapseMatrixType` variants so you just chose :attr:`.SynapseMatrixType.SPARSE`, :attr:`.SynapseMatrixType.DENSE`, :attr:`.SynapseMatrixType.TOEPLITZ` or :attr:`.SynapseMatrixType.PROCEDURAL` (with :attr:`.SynapseMatrixType.DENSE_PROCEDURALG` and :attr:`.SynapseMatrixType.PROCEDURAL_KERNELG` for more unusual options)
-- Extra global parameters only support the 'pointer' form, awaiting a PR to implement settable parameters to replace the other sort
-- Removed implicit variable referneces
\ No newline at end of file
+- Extra global parameters used to be creatable with pointer types to allow models to access
+  arbitrary sized arrays and non-pointer types to create parameters that could be modified at runtime.
+  The latter use case is now handled much more efficiently using the `Dynamic parameters`_ system.
+- Weight update and postsynaptic models could previous refer to neuron variables via _implicit_
+  references. For example ``$(V_pre)`` would refer to a variable called ``V`` on the presynaptic
+  neuron population. This mechanism has been replaced by adding _explicit_ references to
+  weight update, postsynaptic and current source model (see `Variables references`_).
\ No newline at end of file
diff --git a/documentation/5/building_networks.html b/documentation/5/building_networks.html
index a531748c7..2aa9370f3 100644
--- a/documentation/5/building_networks.html
+++ b/documentation/5/building_networks.html
@@ -56,7 +56,7 @@ <h1>Building networks<a class="headerlink" href="#building-networks" title="Perm
 <section id="the-model">
 <h2>The model<a class="headerlink" href="#the-model" title="Permalink to this heading">¶</a></h2>
 <p>A network model is defined as follows:
-A <a class="reference internal" href="source/pygenn.html#pygenn.GeNNModel" title="pygenn.GeNNModel"><code class="xref py py-class docutils literal notranslate"><span class="pre">GeNNModel</span></code></a> must be created with a name and a default precision (see ref floatPrecision)}:</p>
+A <a class="reference internal" href="source/pygenn.html#pygenn.GeNNModel" title="pygenn.GeNNModel"><code class="xref py py-class docutils literal notranslate"><span class="pre">GeNNModel</span></code></a> must be created with a default precision (see ref floatPrecision) and a name:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">model</span> <span class="o">=</span> <span class="n">GeNNModel</span><span class="p">(</span><span class="s2">&quot;float&quot;</span><span class="p">,</span> <span class="s2">&quot;YourModelName&quot;</span><span class="p">)</span>
 </pre></div>
 </div>
@@ -67,7 +67,7 @@ <h2>The model<a class="headerlink" href="#the-model" title="Permalink to this he
                   <span class="n">backend</span><span class="o">=</span><span class="s2">&quot;single_threaded_cpu&quot;</span><span class="p">)</span>
 </pre></div>
 </div>
-<p>When running models on a GPU, smaller models may not fully occupy the device. In some scenerios such as gradient-based training and parameter sweeping,
+<p>When running models on a GPU, smaller models may not fully occupy the device. In some scenarios such as gradient-based training and parameter sweeping,
 this can be overcome by runing multiple copies of the same model at the same time (batching in Machine Learning speak).
 Batching can be enabled on a GeNN model with:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">model</span><span class="o">.</span><span class="n">batch_size</span> <span class="o">=</span> <span class="mi">512</span>
@@ -77,19 +77,25 @@ <h2>The model<a class="headerlink" href="#the-model" title="Permalink to this he
 Whether state variables are duplicated or shared is controlled by the <a class="reference internal" href="source/pygenn.html#pygenn.VarAccess" title="pygenn.VarAccess"><code class="xref py py-class docutils literal notranslate"><span class="pre">VarAccess</span></code></a> or <a class="reference internal" href="source/pygenn.html#pygenn.CustomUpdateVarAccess" title="pygenn.CustomUpdateVarAccess"><code class="xref py py-class docutils literal notranslate"><span class="pre">CustomUpdateVarAccess</span></code></a>
 associated with each variable. Please see TODO for more details.</p>
 <p>Additionally, any preferences exposed by the backend can be configured here.
-For example, the CUDA backend allows you to select which CUDA device to use via the manual_device_id</p>
+For example, the CUDA backend allows you to select which CUDA device to use via the manual_device_id:</p>
+<div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">model</span> <span class="o">=</span> <span class="n">GeNNModel</span><span class="p">(</span><span class="s2">&quot;float&quot;</span><span class="p">,</span> <span class="s2">&quot;YourModelName&quot;</span><span class="p">,</span>
+                 <span class="n">backend</span><span class="o">=</span><span class="s2">&quot;cuda&quot;</span><span class="p">,</span> <span class="n">manual_device_id</span><span class="o">=</span><span class="mi">0</span><span class="p">)</span>
+</pre></div>
+</div>
 </section>
 <section id="populations">
 <h2>Populations<a class="headerlink" href="#populations" title="Permalink to this heading">¶</a></h2>
+<p>Populations formalise the concept of groups of neurons or synapses that are functionally related or a practical grouping, e.g. a brain region in a neuroscience model or a layer in a machine learning context.</p>
 <section id="parameters">
 <h3>Parameters<a class="headerlink" href="#parameters" title="Permalink to this heading">¶</a></h3>
 <p>Parameters are initialised to constant numeric values which are homogeneous across an entire population:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">ini</span> <span class="o">=</span> <span class="p">{</span><span class="s2">&quot;m&quot;</span><span class="p">:</span> <span class="mf">0.0529324</span><span class="p">,</span> <span class="o">...</span><span class="p">}</span>
 </pre></div>
 </div>
-<p>They are very efficient to access from models as their values are either hard-coded into the kernels
+<p>They are very efficient to access from models as their values are either hard-coded into the backend code
 or, on the GPU, delivered via high-performance constant cache.
-However</p>
+However, they can only be used if literally no changes are needed for the entire simulation and all members of
+the population have the exact same parameter value.</p>
 </section>
 <section id="extra-global-parameters">
 <h3>Extra global parameters<a class="headerlink" href="#extra-global-parameters" title="Permalink to this heading">¶</a></h3>
@@ -101,16 +107,16 @@ <h3>Extra global parameter references<a class="headerlink" href="#extra-global-p
 </section>
 <section id="variables">
 <h3>Variables<a class="headerlink" href="#variables" title="Permalink to this heading">¶</a></h3>
-<p>Variables can be initialised in many ways.
-By using the GPU to fill them with a constant value:</p>
+<p>Variables contain values that are individual to the members of a population and can change over time. They can be initialised in many ways. The initialisation is configured through a Python dictionary that is then passed to <a class="reference internal" href="source/pygenn.html#pygenn.GeNNModel.add_neuron_population" title="pygenn.GeNNModel.add_neuron_population"><code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.add_neuron_population()</span></code></a> or <a class="reference internal" href="source/pygenn.html#pygenn.GeNNModel.add_synapse_population" title="pygenn.GeNNModel.add_synapse_population"><code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.add_synapse_population()</span></code></a> which create the populations.</p>
+<p>To initialise variables one can use the backend, e.g. GPU, to fill them with a constant value:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">ini</span> <span class="o">=</span> <span class="p">{</span><span class="s2">&quot;m&quot;</span><span class="p">:</span> <span class="mf">0.0529324</span><span class="p">,</span> <span class="o">...</span><span class="p">}</span>
 </pre></div>
 </div>
-<p>by copying in a sequence of values from Python:</p>
+<p>or copy a sequence of values from Python:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">ini</span> <span class="o">=</span> <span class="p">{</span><span class="s2">&quot;m&quot;</span><span class="p">:</span> <span class="n">np</span><span class="o">.</span><span class="n">arange</span><span class="p">(</span><span class="mf">400.0</span><span class="p">),</span> <span class="o">...</span><span class="p">}</span>
 </pre></div>
 </div>
-<p>or by using a variable initialisation snippet configured using the following function:</p>
+<p>or use a variable initialisation snippet returned by the following function:</p>
 <dl class="py function">
 <dt class="sig sig-object py">
 <span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">init_var</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">snippet</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em><span class="sig-paren">)</span></dt>
@@ -119,28 +125,28 @@ <h3>Variables<a class="headerlink" href="#variables" title="Permalink to this he
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>snippet</strong> (<em>Union</em><em>[</em><em>InitVarSnippetBase</em><em>, </em><em>str</em><em>]</em>) – variable init snippet, either as a string referencing
-a built in snipept (see <a class="reference internal" href="source/pygenn.html#module-pygenn.init_var_snippets" title="pygenn.init_var_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_var_snippets</span></code></a>)
+a built-in snippet (see <a class="reference internal" href="source/pygenn.html#module-pygenn.init_var_snippets" title="pygenn.init_var_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_var_snippets</span></code></a>)
 or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">InitVarSnippetBase</span></code>
 (for example returned by <a class="reference internal" href="source/pygenn.html#pygenn.create_var_init_snippet" title="pygenn.create_var_init_snippet"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_var_init_snippet()</span></code></a>)</p></li>
 <li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the variable init snippet (see <a class="reference internal" href="#parameters">Parameters</a>)</p></li>
 </ul>
 </dd>
 </dl>
-<p>For example, the built in model could be initialised to initialise a variable
+<p>For example, the built-in model “Normal” could be used to initialise a variable
 by sampling from the normal distribution with a mean of 0 and a standard deviation of 1:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">init</span> <span class="o">=</span> <span class="n">init_var</span><span class="p">(</span><span class="s2">&quot;Normal&quot;</span><span class="p">,</span> <span class="p">{</span><span class="s2">&quot;mean&quot;</span><span class="p">:</span> <span class="mf">0.0</span><span class="p">,</span> <span class="s2">&quot;sd&quot;</span><span class="p">:</span> <span class="mf">1.0</span><span class="p">})</span>
 </pre></div>
 </div>
 </dd></dl>
 
-<p>and then used in the dictionary the same way:</p>
+<p>The resulting initialisation snippet can then be used in the dictionary in the usual way:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">ini</span> <span class="o">=</span> <span class="p">{</span><span class="s2">&quot;m&quot;</span><span class="p">:</span> <span class="n">init</span><span class="p">,</span> <span class="o">...</span><span class="p">}</span>
 </pre></div>
 </div>
 </section>
 <section id="variables-references">
 <h3>Variables references<a class="headerlink" href="#variables-references" title="Permalink to this heading">¶</a></h3>
-<p>As well as variables and parameters, various types of model have variable references which are used to reference variables belonging to other populations.
+<p>As well as variables and parameters, various types of models have variable references which are used to reference variables belonging to other populations.
 For example, postsynaptic update models can reference variables in the postsynaptic neuron model and custom updates are ‘attached’ to other populations based on their variable references.</p>
 <p>A variable reference called R could be assigned to various types of variable using the following syntax:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">neuron_var_ref</span> <span class="o">=</span>  <span class="p">{</span><span class="s2">&quot;R&quot;</span><span class="p">:</span> <span class="n">pygenn</span><span class="o">.</span><span class="n">create_var_ref</span><span class="p">(</span><span class="n">ng</span><span class="p">,</span> <span class="s2">&quot;V&quot;</span><span class="p">)}</span>
@@ -203,7 +209,7 @@ <h3>Neuron populations<a class="headerlink" href="#neuron-populations" title="Pe
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>pop_name</strong> (<em>str</em>) – unique name</p></li>
 <li><p><strong>num_neurons</strong> (<em>int</em>) – number of neurons</p></li>
-<li><p><strong>neuron</strong> (<em>Union</em><em>[</em><em>NeuronModelBase</em><em>, </em><em>str</em><em>]</em>) – neuron model either as a string referencing a built in model
+<li><p><strong>neuron</strong> (<em>Union</em><em>[</em><em>NeuronModelBase</em><em>, </em><em>str</em><em>]</em>) – neuron model either as a string referencing a built-in model
 (see <a class="reference internal" href="source/pygenn.html#module-pygenn.neuron_models" title="pygenn.neuron_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">neuron_models</span></code></a>) or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">NeuronModelBase</span></code>
 (for example returned by <a class="reference internal" href="source/pygenn.html#pygenn.create_neuron_model" title="pygenn.create_neuron_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_neuron_model()</span></code></a>)</p></li>
 <li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the neuron model (see <a class="reference internal" href="#parameters">Parameters</a>)</p></li>
@@ -215,7 +221,7 @@ <h3>Neuron populations<a class="headerlink" href="#neuron-populations" title="Pe
 <dd class="field-even"><p><a class="reference internal" href="source/pygenn.html#pygenn.NeuronGroup" title="pygenn._genn.NeuronGroup"><em>NeuronGroup</em></a></p>
 </dd>
 </dl>
-<p>For example, a population of 10 neurons using the built in Izhikevich model and
+<p>For example, a population of 10 neurons using the built-in Izhikevich model and
 the standard set of ‘tonic spiking’ parameters could be added to a model as follows:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">pop</span> <span class="o">=</span> <span class="n">model</span><span class="o">.</span><span class="n">add_neuron_population</span><span class="p">(</span><span class="s2">&quot;pop&quot;</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="s2">&quot;Izhikevich&quot;</span><span class="p">,</span>
                                   <span class="p">{</span><span class="s2">&quot;a&quot;</span><span class="p">:</span> <span class="mf">0.02</span><span class="p">,</span> <span class="s2">&quot;b&quot;</span><span class="p">:</span> <span class="mf">0.2</span><span class="p">,</span> <span class="s2">&quot;c&quot;</span><span class="p">:</span> <span class="o">-</span><span class="mf">65.0</span><span class="p">,</span> <span class="s2">&quot;d&quot;</span><span class="p">:</span> <span class="mf">6.0</span><span class="p">},</span>
@@ -240,7 +246,7 @@ <h3>Synapse populations<a class="headerlink" href="#synapse-populations" title="
 <dl class="field-list simple">
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
-<li><p><strong>snippet</strong> – weight update model either as a string referencing a built in model
+<li><p><strong>snippet</strong> – weight update model either as a string referencing a built-in model
 (see <a class="reference internal" href="source/pygenn.html#module-pygenn.weight_update_models" title="pygenn.weight_update_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">weight_update_models</span></code></a>) or an instance of
 <code class="xref py py-class docutils literal notranslate"><span class="pre">WeightUpdateModelBase</span></code> (for example returned
 by <a class="reference internal" href="source/pygenn.html#pygenn.create_weight_update_model" title="pygenn.create_weight_update_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_weight_update_model()</span></code></a>)</p></li>
@@ -260,7 +266,7 @@ <h3>Synapse populations<a class="headerlink" href="#synapse-populations" title="
 </ul>
 </dd>
 </dl>
-<p>For example, the built in static pulse model with
+<p>For example, the built-in static pulse model with
 constant weights could be initialised as follows:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">weight_init</span> <span class="o">=</span> <span class="n">init_weight_update</span><span class="p">(</span><span class="s2">&quot;StaticPulseConstantWeight&quot;</span><span class="p">,</span> <span class="p">{</span><span class="s2">&quot;g&quot;</span><span class="p">:</span> <span class="mf">1.0</span><span class="p">})</span>
 </pre></div>
@@ -277,7 +283,7 @@ <h3>Synapse populations<a class="headerlink" href="#synapse-populations" title="
 <dl class="field-list simple">
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
-<li><p><strong>snippet</strong> (<em>Union</em><em>[</em><em>PostsynapticModelBase</em><em>, </em><em>str</em><em>]</em>) – postsynaptic model either as a string referencing a built in model
+<li><p><strong>snippet</strong> (<em>Union</em><em>[</em><em>PostsynapticModelBase</em><em>, </em><em>str</em><em>]</em>) – postsynaptic model either as a string referencing a built-in model
 (see <code class="xref py py-mod docutils literal notranslate"><span class="pre">postsynaptic_models_models</span></code>) or an instance of
 <code class="xref py py-class docutils literal notranslate"><span class="pre">PostsynapticModelBase</span></code> (for example returned
 by <a class="reference internal" href="source/pygenn.html#pygenn.create_postsynaptic_model" title="pygenn.create_postsynaptic_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_postsynaptic_model()</span></code></a>)</p></li>
@@ -290,7 +296,7 @@ <h3>Synapse populations<a class="headerlink" href="#synapse-populations" title="
 </ul>
 </dd>
 </dl>
-<p>For example, the built in conductance model with exponential
+<p>For example, the built-in conductance model with exponential
 current shaping could be initialised as follows:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">postsynaptic_init</span> <span class="o">=</span> <span class="n">init_postsynaptic</span><span class="p">(</span><span class="s2">&quot;ExpCond&quot;</span><span class="p">,</span> <span class="p">{</span><span class="s2">&quot;tau&quot;</span><span class="p">:</span> <span class="mf">1.0</span><span class="p">,</span> <span class="s2">&quot;E&quot;</span><span class="p">:</span> <span class="o">-</span><span class="mf">80.0</span><span class="p">},</span>
                                       <span class="n">var_refs</span><span class="o">=</span><span class="p">{</span><span class="s2">&quot;V&quot;</span><span class="p">:</span> <span class="n">create_var_ref</span><span class="p">(</span><span class="n">pop1</span><span class="p">,</span> <span class="s2">&quot;V&quot;</span><span class="p">)})</span>
@@ -327,14 +333,14 @@ <h3>Synapse populations<a class="headerlink" href="#synapse-populations" title="
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>snippet</strong> (<em>Union</em><em>[</em><em>InitSparseConnectivitySnippetBase</em><em>, </em><em>str</em><em>]</em>) – sparse connectivity init snippet, either as a string referencing
-a built in snipept (see <a class="reference internal" href="source/pygenn.html#module-pygenn.init_sparse_connectivity_snippets" title="pygenn.init_sparse_connectivity_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_sparse_connectivity_snippets</span></code></a>)
+a built-in snippet (see <a class="reference internal" href="source/pygenn.html#module-pygenn.init_sparse_connectivity_snippets" title="pygenn.init_sparse_connectivity_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_sparse_connectivity_snippets</span></code></a>)
 or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">InitSparseConnectivitySnippetBase</span></code>
 (for example returned by <a class="reference internal" href="source/pygenn.html#pygenn.create_sparse_connect_init_snippet" title="pygenn.create_sparse_connect_init_snippet"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_sparse_connect_init_snippet()</span></code></a>)</p></li>
 <li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the sparse connectivity init snippet (see <a class="reference internal" href="#parameters">Parameters</a>)</p></li>
 </ul>
 </dd>
 </dl>
-<p>For example, the built in model could be initialised to generate connectivity
+<p>For example, the built-in model could be initialised to generate connectivity
 where each pair of pre and postsynaptic neurons is connected with a probability of 0.1:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">init</span> <span class="o">=</span> <span class="n">init_sparse_connectivity</span><span class="p">(</span><span class="s2">&quot;FixedProbability&quot;</span><span class="p">,</span> <span class="p">{</span><span class="s2">&quot;prob&quot;</span><span class="p">:</span> <span class="mf">0.1</span><span class="p">})</span>
 </pre></div>
@@ -351,14 +357,14 @@ <h3>Synapse populations<a class="headerlink" href="#synapse-populations" title="
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>snippet</strong> – toeplitz connectivity init snippet, either as a string referencing
-a built in snipept (see <a class="reference internal" href="source/pygenn.html#module-pygenn.init_toeplitz_connectivity_snippets" title="pygenn.init_toeplitz_connectivity_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_toeplitz_connectivity_snippets</span></code></a>)
+a built-in snippet (see <a class="reference internal" href="source/pygenn.html#module-pygenn.init_toeplitz_connectivity_snippets" title="pygenn.init_toeplitz_connectivity_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_toeplitz_connectivity_snippets</span></code></a>)
 or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">InitToeplitzConnectivitySnippetBase</span></code>
 (for example returned by <a class="reference internal" href="source/pygenn.html#pygenn.create_toeplitz_connect_init_snippet" title="pygenn.create_toeplitz_connect_init_snippet"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_toeplitz_connect_init_snippet()</span></code></a>)</p></li>
 <li><p><strong>params</strong> – parameter values for the toeplitz connectivity init snippet (see <a class="reference internal" href="#parameters">Parameters</a>)</p></li>
 </ul>
 </dd>
 </dl>
-<p>For example, the built in model could be initialised to generate 2D convolutional
+<p>For example, the built-in model could be initialised to generate 2D convolutional
 connectivity with a <span class="math notranslate nohighlight">\(3 \times 3\)</span> kernel, a <span class="math notranslate nohighlight">\(64 \times 64 \times 1\)</span> input
 and a <span class="math notranslate nohighlight">\(62 \times 62 \times 1\)</span> input:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">params</span> <span class="o">=</span> <span class="p">{</span><span class="s2">&quot;conv_kh&quot;</span><span class="p">:</span> <span class="mi">3</span><span class="p">,</span> <span class="s2">&quot;conv_kw&quot;</span><span class="p">:</span> <span class="mi">3</span><span class="p">,</span>
@@ -427,7 +433,7 @@ <h3>Current sources<a class="headerlink" href="#current-sources" title="Permalin
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>cs_name</strong> (<em>str</em>) – unique name</p></li>
-<li><p><strong>current_source_model</strong> (<em>Union</em><em>[</em><em>CurrentSourceModelBase</em><em>, </em><em>str</em><em>]</em>) – current source model either as a string referencing a built in model
+<li><p><strong>current_source_model</strong> (<em>Union</em><em>[</em><em>CurrentSourceModelBase</em><em>, </em><em>str</em><em>]</em>) – current source model either as a string referencing a built-in model
 (see <a class="reference internal" href="source/pygenn.html#module-pygenn.current_source_models" title="pygenn.current_source_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">current_source_models</span></code></a>) or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">CurrentSourceModelBase</span></code>
 (for example returned by <a class="reference internal" href="source/pygenn.html#pygenn.create_current_source_model" title="pygenn.create_current_source_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_current_source_model()</span></code></a>)</p></li>
 <li><p><strong>pop</strong> (<a class="reference internal" href="source/pygenn.html#pygenn.NeuronGroup" title="pygenn._genn.NeuronGroup"><em>NeuronGroup</em></a>) – neuron population to inject current into</p></li>
@@ -472,7 +478,7 @@ <h3>Custom updates<a class="headerlink" href="#custom-updates" title="Permalink
 <li><p><strong>cu_name</strong> (<em>str</em>) – unique name</p></li>
 <li><p><strong>group_name</strong> (<em>str</em>) – name of the ‘custom update group’ to include this update in.
 All custom updates in the same group are executed simultaneously.</p></li>
-<li><p><strong>custom_update_model</strong> (<em>Union</em><em>[</em><em>CustomUpdateModelBase</em><em>, </em><em>str</em><em>]</em>) – custom update model either as a string referencing a built in model
+<li><p><strong>custom_update_model</strong> (<em>Union</em><em>[</em><em>CustomUpdateModelBase</em><em>, </em><em>str</em><em>]</em>) – custom update model either as a string referencing a built-in model
 (see <a class="reference internal" href="source/pygenn.html#module-pygenn.custom_update_models" title="pygenn.custom_update_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">custom_update_models</span></code></a>) or an instance of
 <code class="xref py py-class docutils literal notranslate"><span class="pre">CustomUpdateModelBase</span></code>
 (for example returned by <a class="reference internal" href="source/pygenn.html#pygenn.create_custom_update_model" title="pygenn.create_custom_update_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_custom_update_model()</span></code></a>)</p></li>
@@ -499,7 +505,7 @@ <h3>Custom connectivity updates<a class="headerlink" href="#custom-connectivity-
 Custom connectivity updates are added to a model using:</p>
 <dl class="py method">
 <dt class="sig sig-object py">
-<span class="sig-prename descclassname"><span class="pre">GeNNModel.</span></span><span class="sig-name descname"><span class="pre">add_custom_connectivity_update</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">cu_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">group_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">syn_group</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_conn_update_model</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em><span class="sig-paren">)</span></dt>
+<span class="sig-prename descclassname"><span class="pre">GeNNModel.</span></span><span class="sig-name descname"><span class="pre">add_custom_connectivity_update</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">cu_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">group_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">syn_group</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_conn_update_model</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">egp_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em><span class="sig-paren">)</span></dt>
 <dd><p>Add a custom connectivity update to the GeNN model</p>
 <dl class="field-list simple">
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
@@ -508,7 +514,7 @@ <h3>Custom connectivity updates<a class="headerlink" href="#custom-connectivity-
 <li><p><strong>group_name</strong> (<em>str</em>) – name of the ‘custom update group’ to include this update in.
 All custom updates in the same group are executed simultaneously.</p></li>
 <li><p><strong>syn_group</strong> (<a class="reference internal" href="source/pygenn.html#pygenn.SynapseGroup" title="pygenn._genn.SynapseGroup"><em>SynapseGroup</em></a>) – Synapse group to attach custom connectivity update to</p></li>
-<li><p><strong>custom_conn_update_model</strong> (<em>Union</em><em>[</em><em>CustomConnectivityUpdateModelBase</em><em>, </em><em>str</em><em>]</em>) – custom connectivity update model either as a string referencing a built in model
+<li><p><strong>custom_conn_update_model</strong> (<em>Union</em><em>[</em><em>CustomConnectivityUpdateModelBase</em><em>, </em><em>str</em><em>]</em>) – custom connectivity update model either as a string referencing a built-in model
 (see <a class="reference internal" href="source/pygenn.html#module-pygenn.custom_connectivity_update_models" title="pygenn.custom_connectivity_update_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">custom_connectivity_update_models</span></code></a>) or an instance of
 <code class="xref py py-class docutils literal notranslate"><span class="pre">CustomConnectivityUpdateModelBaseUpdateModelBase</span></code>
 (for example returned by <a class="reference internal" href="source/pygenn.html#pygenn.create_custom_connectivity_update_model" title="pygenn.create_custom_connectivity_update_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_custom_connectivity_update_model()</span></code></a>)</p></li>
@@ -528,6 +534,9 @@ <h3>Custom connectivity updates<a class="headerlink" href="#custom-connectivity-
 <li><p><strong>post_var_refs</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>VarReference</em><em>]</em>) – references to postsynaptic variables,
 typically created using <a class="reference internal" href="source/pygenn.html#pygenn.create_var_ref" title="pygenn.create_var_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_var_ref()</span></code></a>
 (see <a class="reference internal" href="#variables-references">Variables references</a>)</p></li>
+<li><p><strong>egp_refs</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>EGPReference</em><em>]</em>) – references to extra global parameters in other populations
+to access from this update, typically created using
+<a class="reference internal" href="source/pygenn.html#pygenn.create_egp_ref" title="pygenn.create_egp_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_egp_ref()</span></code></a> (see <a class="reference internal" href="#extra-global-parameter-references">Extra global parameter references</a>).</p></li>
 </ul>
 </dd>
 </dl>
diff --git a/documentation/5/custom_models.html b/documentation/5/custom_models.html
index 994165aaf..cb48dbe21 100644
--- a/documentation/5/custom_models.html
+++ b/documentation/5/custom_models.html
@@ -109,7 +109,7 @@ <h3>Variable initialisation<a class="headerlink" href="#variable-initialisation"
 <dd><p>Creates a new variable initialisation snippet
 Within the <code class="docutils literal notranslate"><span class="pre">var_init_code</span></code>, the parameters, derived parameters and
 extra global parameters defined in this snippet can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 </ul>
@@ -172,7 +172,7 @@ <h3>Sparse connectivity initialisation<a class="headerlink" href="#sparse-connec
 <dd><p>Creates a new sparse connectivity initialisation snippet
 Within the code strings, the parameters, derived parameters and
 extra global parameters defined in this snippet can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">num_pre</span></code> which represents the number of presynaptic neurons</p></li>
@@ -180,14 +180,16 @@ <h3>Sparse connectivity initialisation<a class="headerlink" href="#sparse-connec
 <li><p><code class="docutils literal notranslate"><span class="pre">thread</span></code> when some procedural connectivity is used with multiple
 threads per presynaptic neuron, represents the index of thread</p></li>
 </ul>
-<p>and, in <code class="docutils literal notranslate"><span class="pre">row_build_code</span></code>:
-- <code class="docutils literal notranslate"><span class="pre">id_pre</span></code> represents the index of the presynaptic neuron (starting from zero)
-- <code class="docutils literal notranslate"><span class="pre">id_post_begin</span></code> when some procedural connectivity is used with multiple</p>
-<blockquote>
-<div><p>threads per presynaptic neuron, represents the index of the first postsynaptic neuron to connect.</p>
-</div></blockquote>
-<p>and, in <code class="docutils literal notranslate"><span class="pre">col_build_code</span></code>:
-- <code class="docutils literal notranslate"><span class="pre">id_post</span></code> which represents the index of the postsynaptic neuron (starting from zero)</p>
+<p>and, in <code class="docutils literal notranslate"><span class="pre">row_build_code</span></code>:</p>
+<ul class="simple">
+<li><p><code class="docutils literal notranslate"><span class="pre">id_pre</span></code> represents the index of the presynaptic neuron (starting from zero)</p></li>
+<li><p><code class="docutils literal notranslate"><span class="pre">id_post_begin</span></code> when some procedural connectivity is used with multiple
+threads per presynaptic neuron, represents the index of the first postsynaptic neuron to connect.</p></li>
+</ul>
+<p>and, in <code class="docutils literal notranslate"><span class="pre">col_build_code</span></code>:</p>
+<ul class="simple">
+<li><p><code class="docutils literal notranslate"><span class="pre">id_post</span></code> which represents the index of the postsynaptic neuron (starting from zero)</p></li>
+</ul>
 <p>Finally, the function <code class="docutils literal notranslate"><span class="pre">addSynapse(x)</span></code> can be used to add a new synapse to the connectivity
 where, in <code class="docutils literal notranslate"><span class="pre">row_build_code</span></code>, <code class="docutils literal notranslate"><span class="pre">x</span></code> is the index of the postsynaptic neuron to connect <cite>id_pre`</cite> to
 and, in <code class="docutils literal notranslate"><span class="pre">col_build_code</span></code>, <code class="docutils literal notranslate"><span class="pre">x</span></code> is the index of the presynaptic neuron to connect to <code class="docutils literal notranslate"><span class="pre">id_post</span></code></p>
@@ -251,7 +253,7 @@ <h3>Toeplitz connectivity initialisation<a class="headerlink" href="#toeplitz-co
 snippet of code specified using the <cite>diagonal_build_code`</cite>.
 Within the code strings, the parameters, derived parameters and
 extra global parameters defined in this snippet can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">num_pre</span></code> which represents the number of presynaptic neurons</p></li>
@@ -326,7 +328,7 @@ <h3>Neuron models<a class="headerlink" href="#neuron-models" title="Permalink to
 Within all of the code strings, the variables, parameters,
 derived parameters, additional input variables and extra global
 parameters defined in this model can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">Isyn</span></code> which represents the total incoming synaptic input.</p></li>
@@ -580,7 +582,7 @@ <h3>Postsynaptic models<a class="headerlink" href="#postsynaptic-models" title="
 Within all of the code strings, the variables, parameters,
 derived parameters and extra global parameters defined in this model
 can all be referred to be name. Additionally, the code may refer to the
-following built in read-only variables:</p>
+following built-in read-only variables:</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">id</span></code> which represents a neurons index within a population (starting from zero)</p></li>
@@ -622,7 +624,7 @@ <h3>Current source models<a class="headerlink" href="#current-source-models" tit
 Within the <code class="docutils literal notranslate"><span class="pre">injection_code</span></code> code string, the variables, parameters,
 derived parameters, neuron variable references and extra global
 parameters defined in this model can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">id</span></code> which represents a neurons index within a population (starting from zero)</p></li>
@@ -672,7 +674,7 @@ <h3>Custom update models<a class="headerlink" href="#custom-update-models" title
 Within the <code class="docutils literal notranslate"><span class="pre">update_code</span></code> code string, the variables, parameters,
 derived parameters, variable references, extra global parameters
 and extra global parameter references defined in this model can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 </ul>
@@ -764,7 +766,7 @@ <h3>Custom connectivity update models<a class="headerlink" href="#custom-connect
 New custom connectivity update models are defined by calling:</p>
 <dl class="py function">
 <dt class="sig sig-object py">
-<span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">create_custom_connectivity_update_model</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">class_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">derived_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">row_update_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">host_update_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">extra_global_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span></dt>
+<span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">create_custom_connectivity_update_model</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">class_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">derived_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">row_update_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">host_update_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">extra_global_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">extra_global_param_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span></dt>
 <dd><p>Creates a new custom connectivity update model.</p>
 <p>Within host update code, you have full access to parameters, derived parameters,
 extra global parameters and pre and postsynaptic variables. By design you do
@@ -800,6 +802,8 @@ <h3>Custom connectivity update models<a class="headerlink" href="#custom-connect
 on CPU when custom connectivity update is launched</p></li>
 <li><p><strong>extra_global_params</strong> – names and types of model
 extra global parameters</p></li>
+<li><p><strong>extra_global_param_refs</strong> – names and types of extra global
+parameter references</p></li>
 <li><p><strong>post_vars</strong> (<em>Optional</em><em>[</em><em>Sequence</em><em>[</em><em>Union</em><em>[</em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>]</em><em>, </em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>, </em><a class="reference internal" href="source/pygenn.html#pygenn.VarAccess" title="pygenn._genn.VarAccess"><em>VarAccess</em></a><em>]</em><em>]</em><em>]</em><em>]</em>) – </p></li>
 <li><p><strong>pre_var_refs</strong> (<em>Optional</em><em>[</em><em>Sequence</em><em>[</em><em>Union</em><em>[</em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>]</em><em>, </em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>, </em><a class="reference internal" href="source/pygenn.html#pygenn.VarAccessMode" title="pygenn._genn.VarAccessMode"><em>VarAccessMode</em></a><em>]</em><em>]</em><em>]</em><em>]</em>) – </p></li>
 <li><p><strong>post_var_refs</strong> (<em>Optional</em><em>[</em><em>Sequence</em><em>[</em><em>Union</em><em>[</em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>]</em><em>, </em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>, </em><a class="reference internal" href="source/pygenn.html#pygenn.VarAccessMode" title="pygenn._genn.VarAccessMode"><em>VarAccessMode</em></a><em>]</em><em>]</em><em>]</em><em>]</em>) – </p></li>
diff --git a/documentation/5/genindex.html b/documentation/5/genindex.html
index 9e5f52a5f..20db564d1 100644
--- a/documentation/5/genindex.html
+++ b/documentation/5/genindex.html
@@ -145,8 +145,12 @@ <h2 id="C">C</h2>
       <li><a href="source/pygenn.html#pygenn.create_egp_ref">create_egp_ref() (in module pygenn)</a>
 </li>
       <li><a href="source/pygenn.html#pygenn.create_neuron_model">create_neuron_model() (in module pygenn)</a>
+</li>
+      <li><a href="source/pygenn.html#pygenn.create_post_var_ref">create_post_var_ref() (in module pygenn)</a>
 </li>
       <li><a href="source/pygenn.html#pygenn.create_postsynaptic_model">create_postsynaptic_model() (in module pygenn)</a>
+</li>
+      <li><a href="source/pygenn.html#pygenn.create_pre_var_ref">create_pre_var_ref() (in module pygenn)</a>
 </li>
       <li><a href="source/pygenn.html#pygenn.create_psm_egp_ref">create_psm_egp_ref() (in module pygenn)</a>
 </li>
@@ -159,11 +163,11 @@ <h2 id="C">C</h2>
       <li><a href="source/pygenn.html#pygenn.create_var_init_snippet">create_var_init_snippet() (in module pygenn)</a>
 </li>
       <li><a href="source/pygenn.html#pygenn.create_var_ref">create_var_ref() (in module pygenn)</a>
-</li>
-      <li><a href="source/pygenn.html#pygenn.create_weight_update_model">create_weight_update_model() (in module pygenn)</a>
 </li>
   </ul></td>
   <td style="width: 33%; vertical-align: top;"><ul>
+      <li><a href="source/pygenn.html#pygenn.create_weight_update_model">create_weight_update_model() (in module pygenn)</a>
+</li>
       <li><a href="source/pygenn.html#pygenn.create_wu_egp_ref">create_wu_egp_ref() (in module pygenn)</a>
 </li>
       <li><a href="source/pygenn.html#pygenn.create_wu_post_var_ref">create_wu_post_var_ref() (in module pygenn)</a>
diff --git a/documentation/5/index.html b/documentation/5/index.html
index 13e2c4c72..e29ce6ae3 100644
--- a/documentation/5/index.html
+++ b/documentation/5/index.html
@@ -22,7 +22,7 @@
     
     <link rel="index" title="Index" href="genindex.html" />
     <link rel="search" title="Search" href="search.html" />
-    <link rel="next" title="Upgrading from GeNN 4" href="upgrading.html" /> 
+    <link rel="next" title="Installation" href="installation.html" /> 
   </head><body>
     <div class="related" role="navigation" aria-label="related navigation">
       <h3>Navigation</h3>
@@ -34,7 +34,7 @@ <h3>Navigation</h3>
           <a href="py-modindex.html" title="Python Module Index"
              >modules</a> |</li>
         <li class="right" >
-          <a href="upgrading.html" title="Upgrading from GeNN 4"
+          <a href="installation.html" title="Installation"
              accesskey="N">next</a> |</li>
         <li class="nav-item nav-item-0"><a href="#">PyGeNN  documentation</a> &#187;</li>
         <li class="nav-item nav-item-this"><a href="">PyGeNN documentation</a></li> 
@@ -56,6 +56,7 @@ <h1>PyGeNN documentation<a class="headerlink" href="#pygenn-documentation" title
 <p>Note, this documentation is under construction. If you cannot find what you are looking for, please contact the project developers.</p>
 <div class="toctree-wrapper compound">
 <ul>
+<li class="toctree-l1"><a class="reference internal" href="installation.html">Installation</a></li>
 <li class="toctree-l1"><a class="reference internal" href="upgrading.html">Upgrading from GeNN 4</a></li>
 <li class="toctree-l1"><a class="reference internal" href="building_networks.html">Building networks</a></li>
 <li class="toctree-l1"><a class="reference internal" href="simulating_networks.html">Simulating networks</a></li>
@@ -103,8 +104,8 @@ <h3><a href="#">Table of Contents</a></h3>
   </div>
   <div>
     <h4>Next topic</h4>
-    <p class="topless"><a href="upgrading.html"
-                          title="next chapter">Upgrading from GeNN 4</a></p>
+    <p class="topless"><a href="installation.html"
+                          title="next chapter">Installation</a></p>
   </div>
   <div role="note" aria-label="source link">
     <h3>This Page</h3>
@@ -137,7 +138,7 @@ <h3>Navigation</h3>
           <a href="py-modindex.html" title="Python Module Index"
              >modules</a> |</li>
         <li class="right" >
-          <a href="upgrading.html" title="Upgrading from GeNN 4"
+          <a href="installation.html" title="Installation"
              >next</a> |</li>
         <li class="nav-item nav-item-0"><a href="#">PyGeNN  documentation</a> &#187;</li>
         <li class="nav-item nav-item-this"><a href="">PyGeNN documentation</a></li> 
diff --git a/documentation/5/installation.html b/documentation/5/installation.html
new file mode 100644
index 000000000..0e44d6156
--- /dev/null
+++ b/documentation/5/installation.html
@@ -0,0 +1,185 @@
+
+<!DOCTYPE html>
+
+<html lang="en">
+  <head>
+    <meta charset="utf-8" />
+    <meta name="viewport" content="width=device-width, initial-scale=1.0" /><meta name="generator" content="Docutils 0.18.1: http://docutils.sourceforge.net/" />
+
+    <title>Installation &#8212; PyGeNN  documentation</title>
+    <link rel="stylesheet" type="text/css" href="_static/pygments.css" />
+    <link rel="stylesheet" type="text/css" href="_static/classic.css" />
+    <link rel="stylesheet" type="text/css" href="_static/sg_gallery.css" />
+    <link rel="stylesheet" type="text/css" href="_static/sg_gallery-binder.css" />
+    <link rel="stylesheet" type="text/css" href="_static/sg_gallery-dataframe.css" />
+    <link rel="stylesheet" type="text/css" href="_static/sg_gallery-rendered-html.css" />
+    
+    <script data-url_root="./" id="documentation_options" src="_static/documentation_options.js"></script>
+    <script src="_static/jquery.js"></script>
+    <script src="_static/underscore.js"></script>
+    <script src="_static/_sphinx_javascript_frameworks_compat.js"></script>
+    <script src="_static/doctools.js"></script>
+    
+    <link rel="index" title="Index" href="genindex.html" />
+    <link rel="search" title="Search" href="search.html" />
+    <link rel="next" title="Upgrading from GeNN 4" href="upgrading.html" />
+    <link rel="prev" title="PyGeNN documentation" href="index.html" /> 
+  </head><body>
+    <div class="related" role="navigation" aria-label="related navigation">
+      <h3>Navigation</h3>
+      <ul>
+        <li class="right" style="margin-right: 10px">
+          <a href="genindex.html" title="General Index"
+             accesskey="I">index</a></li>
+        <li class="right" >
+          <a href="py-modindex.html" title="Python Module Index"
+             >modules</a> |</li>
+        <li class="right" >
+          <a href="upgrading.html" title="Upgrading from GeNN 4"
+             accesskey="N">next</a> |</li>
+        <li class="right" >
+          <a href="index.html" title="PyGeNN documentation"
+             accesskey="P">previous</a> |</li>
+        <li class="nav-item nav-item-0"><a href="index.html">PyGeNN  documentation</a> &#187;</li>
+        <li class="nav-item nav-item-this"><a href="">Installation</a></li> 
+      </ul>
+    </div>  
+
+    <div class="document">
+      <div class="documentwrapper">
+        <div class="bodywrapper">
+          <div class="body" role="main">
+            
+  <section id="installation">
+<h1>Installation<a class="headerlink" href="#installation" title="Permalink to this heading">¶</a></h1>
+<p>In future we plan on providing binary builds of GeNN via conda. However, for now, GeNN
+needs to be installed from source.</p>
+<section id="pre-installation">
+<h2>Pre-installation<a class="headerlink" href="#pre-installation" title="Permalink to this heading">¶</a></h2>
+<ol class="arabic simple">
+<li><p>Install the C++ compiler on the machine, if not already present.
+For Windows, Visual Studio 2019 or above is required. The Microsoft Visual Studio
+Community Edition can be downloaded from
+<a class="reference external" href="https://www.visualstudio.com/en-us/downloads/download-visual-studio-vs.aspx">https://www.visualstudio.com/en-us/downloads/download-visual-studio-vs.aspx</a>.
+When installing Visual Studio, one should select the ‘Desktop
+development with C++’ configuration. On Linux, the GNU Compiler
+Collection (GCC) 7.5 or above is required. This can be obtained from your
+Linux distribution repository, for example on Ubuntu by running <code class="docutils literal notranslate"><span class="pre">sudo</span> <span class="pre">apt-get</span> <span class="pre">install</span> <span class="pre">g++</span></code>,
+or alternatively from <a class="reference external" href="https://gcc.gnu.org/index.html">https://gcc.gnu.org/index.html</a>.</p></li>
+<li><p>If your machine has an NVIDIA GPU and you haven’t installed CUDA already,
+obtain a fresh installation of the NVIDIA CUDA toolkit from
+<a class="reference external" href="https://developer.nvidia.com/cuda-downloads">https://developer.nvidia.com/cuda-downloads</a>
+Again, be sure to pick CUDA and C++ compiler versions which are compatible
+with each other. The latest C++ compiler need not necessarily be
+compatible with the latest CUDA toolkit.</p></li>
+<li><p>GeNN uses the <code class="docutils literal notranslate"><span class="pre">CUDA_PATH</span></code> environment variable to determine which
+version of CUDA to build against. On Windows, this is set automatically when
+installing CUDA. However, if you choose, you can verify which version is
+selected by running <code class="docutils literal notranslate"><span class="pre">echo</span> <span class="pre">%CUDA_PATH</span></code> in a command prompt.
+However, on Linux and Mac you need to set <code class="docutils literal notranslate"><span class="pre">CUDA_PATH</span></code> manually with:
+<code class="docutils literal notranslate"><span class="pre">`bash</span>
+<span class="pre">export</span> <span class="pre">CUDA_PATH=/usr/local/cuda</span>
+<span class="pre">`</span></code>
+assuming CUDA is installed in /usr/local/cuda (the standard location
+on Ubuntu Linux). Again, to make this change persistent, this can
+be added to your login script (e.g. <code class="docutils literal notranslate"><span class="pre">.profile</span></code> or <code class="docutils literal notranslate"><span class="pre">.bashrc</span></code>)</p></li>
+<li><p>Either download the latest release of GeNN and extract into your
+home directory or clone using git from <a class="reference external" href="https://github.com/genn-team/genn">https://github.com/genn-team/genn</a></p></li>
+<li><p>On Linux, install the development version of libffi. For example, on Ubuntu you can do this
+by running <code class="docutils literal notranslate"><span class="pre">sudo</span> <span class="pre">apt-get</span> <span class="pre">install</span> <span class="pre">libffi-dev</span></code>.</p></li>
+<li><p>Install the pybind11, psutil and numpy packages with pip i.e. <code class="docutils literal notranslate"><span class="pre">pip</span> <span class="pre">install</span> <span class="pre">pybind11</span> <span class="pre">psutil</span> <span class="pre">numpy</span></code>.</p></li>
+</ol>
+</section>
+<section id="building-with-setup-py">
+<h2>Building with setup.py<a class="headerlink" href="#building-with-setup-py" title="Permalink to this heading">¶</a></h2>
+<p>From the GeNN directory, the GeNN libraries and python package can be built
+with <code class="docutils literal notranslate"><span class="pre">python</span> <span class="pre">setup.py</span> <span class="pre">install</span></code>. If you wish to create an editable install
+(most useful if you are intending to modify GeNN yourself) you can also used
+<code class="docutils literal notranslate"><span class="pre">python</span> <span class="pre">setup.py</span> <span class="pre">develop</span></code>. On Linux (or Windows if you have a debug version
+of the python libraries installed) you can build a debug version of GeNN with
+<code class="docutils literal notranslate"><span class="pre">python</span> <span class="pre">setup.py</span> <span class="pre">build_ext</span> <span class="pre">--debug</span> <span class="pre">develop</span></code>.</p>
+</section>
+<section id="building-with-pip">
+<h2>Building with pip<a class="headerlink" href="#building-with-pip" title="Permalink to this heading">¶</a></h2>
+<p>From the GeNN directory, the GeNN libraries and python package can be built
+with <code class="docutils literal notranslate"><span class="pre">pip</span> <span class="pre">install</span> <span class="pre">.</span></code>. If you wish to create an editable install
+(most useful if you are intending to modify GeNN yourself) you can also used
+<code class="docutils literal notranslate"><span class="pre">pip</span> <span class="pre">install</span> <span class="pre">--editable</span> <span class="pre">.</span></code>.</p>
+</section>
+</section>
+
+
+            <div class="clearer"></div>
+          </div>
+        </div>
+      </div>
+      <div class="sphinxsidebar" role="navigation" aria-label="main navigation">
+        <div class="sphinxsidebarwrapper">
+  <div>
+    <h3><a href="index.html">Table of Contents</a></h3>
+    <ul>
+<li><a class="reference internal" href="#">Installation</a><ul>
+<li><a class="reference internal" href="#pre-installation">Pre-installation</a></li>
+<li><a class="reference internal" href="#building-with-setup-py">Building with setup.py</a></li>
+<li><a class="reference internal" href="#building-with-pip">Building with pip</a></li>
+</ul>
+</li>
+</ul>
+
+  </div>
+  <div>
+    <h4>Previous topic</h4>
+    <p class="topless"><a href="index.html"
+                          title="previous chapter">PyGeNN documentation</a></p>
+  </div>
+  <div>
+    <h4>Next topic</h4>
+    <p class="topless"><a href="upgrading.html"
+                          title="next chapter">Upgrading from GeNN 4</a></p>
+  </div>
+  <div role="note" aria-label="source link">
+    <h3>This Page</h3>
+    <ul class="this-page-menu">
+      <li><a href="_sources/installation.rst.txt"
+            rel="nofollow">Show Source</a></li>
+    </ul>
+   </div>
+<div id="searchbox" style="display: none" role="search">
+  <h3 id="searchlabel">Quick search</h3>
+    <div class="searchformwrapper">
+    <form class="search" action="search.html" method="get">
+      <input type="text" name="q" aria-labelledby="searchlabel" autocomplete="off" autocorrect="off" autocapitalize="off" spellcheck="false"/>
+      <input type="submit" value="Go" />
+    </form>
+    </div>
+</div>
+<script>document.getElementById('searchbox').style.display = "block"</script>
+        </div>
+      </div>
+      <div class="clearer"></div>
+    </div>
+    <div class="related" role="navigation" aria-label="related navigation">
+      <h3>Navigation</h3>
+      <ul>
+        <li class="right" style="margin-right: 10px">
+          <a href="genindex.html" title="General Index"
+             >index</a></li>
+        <li class="right" >
+          <a href="py-modindex.html" title="Python Module Index"
+             >modules</a> |</li>
+        <li class="right" >
+          <a href="upgrading.html" title="Upgrading from GeNN 4"
+             >next</a> |</li>
+        <li class="right" >
+          <a href="index.html" title="PyGeNN documentation"
+             >previous</a> |</li>
+        <li class="nav-item nav-item-0"><a href="index.html">PyGeNN  documentation</a> &#187;</li>
+        <li class="nav-item nav-item-this"><a href="">Installation</a></li> 
+      </ul>
+    </div>
+    <div class="footer" role="contentinfo">
+        &#169; Copyright 2024, James Knight, Thomas Nowotny.
+      Created using <a href="https://www.sphinx-doc.org/">Sphinx</a> 5.0.2.
+    </div>
+  </body>
+</html>
\ No newline at end of file
diff --git a/documentation/5/objects.inv b/documentation/5/objects.inv
index 5727466f4..aaa9f793a 100644
Binary files a/documentation/5/objects.inv and b/documentation/5/objects.inv differ
diff --git a/documentation/5/searchindex.js b/documentation/5/searchindex.js
index 17587deda..cb0473c1f 100644
--- a/documentation/5/searchindex.js
+++ b/documentation/5/searchindex.js
@@ -1 +1 @@
-Search.setIndex({"docnames": ["bibliography", "building_networks", "custom_models", "index", "sg_execution_times", "simulating_networks", "source/modules", "source/pygenn", "upgrading", "userproject/index", "userproject/mnist_mb_classifier", "userproject/potjans_microcircuit", "userproject/sg_execution_times", "userproject/superspike_demo"], "filenames": ["bibliography.rst", "building_networks.rst", "custom_models.rst", "index.rst", "sg_execution_times.rst", "simulating_networks.rst", "source\\modules.rst", "source\\pygenn.rst", "upgrading.rst", "userproject\\index.rst", "userproject\\mnist_mb_classifier.rst", "userproject\\potjans_microcircuit.rst", "userproject\\sg_execution_times.rst", "userproject\\superspike_demo.rst"], "titles": ["Bibliography", "Building networks", "Custom models", "PyGeNN documentation", "Computation times", "Simulating networks", "pygenn", "pygenn package", "Upgrading from GeNN 4", "User projects", "MNIST classification using an insect-inspired mushroom body model", "PyGeNN implementation of local cortical microcircuit model", "Computation times", "PyGeNN implementation of SuperSpike"], "terms": {"morrison2008": [0, 2, 7], "morrison": 0, "A": [0, 1, 2, 7], "diesmann": [0, 11], "m": [0, 1, 7, 13], "gerstner": 0, "w": [0, 2, 13], "2008": 0, "phenomenolog": 0, "model": [0, 3, 4, 5, 7, 8, 9, 12, 13], "synapt": [0, 1, 2, 7, 10, 11, 13], "plastic": 0, "base": [0, 1, 2, 5, 7, 13], "spike": [0, 1, 2, 3, 7, 10, 11, 13], "time": [0, 1, 2, 5, 7, 8, 10, 11, 13], "biolog": 0, "cybernet": 0, "98": [0, 7], "459": 0, "478": 0, "http": [0, 2], "doi": 0, "org": 0, "10": [0, 1, 5, 7, 10, 11, 13], "1007": 0, "s00422": 0, "008": [0, 11], "0233": [0, 10], "1": [0, 1, 2, 3, 5, 7, 10, 11, 13], "potjans2014": [0, 11], "potjan": [0, 11], "t": [0, 2, 5, 7, 8, 10, 11, 13], "c": [0, 1, 2, 3, 7, 10, 11, 13], "2014": 0, "The": [0, 2, 3, 5, 7, 8, 11], "cell": [0, 10, 11], "type": [0, 1, 2, 6, 11, 13], "specif": [0, 11], "cortic": [0, 4, 9, 12], "microcircuit": [0, 4, 9, 12], "relat": [0, 7], "structur": [0, 1, 2, 8, 13], "activ": [0, 2, 7, 11], "full": [0, 2, 7, 10, 11, 13], "scale": [0, 7, 10, 11], "network": [0, 2, 3, 7, 10, 11, 13], "cerebr": 0, "cortex": 0, "24": 0, "3": [0, 1, 4, 7, 10, 11, 12, 13], "785": 0, "806": 0, "1093": 0, "cercor": 0, "bhs358": 0, "zenke2018": [0, 13], "zenk": [0, 13], "f": [0, 10, 11, 13], "ganguli": [0, 13], "s": [0, 1, 2, 5, 7, 10, 11, 13], "2018": 0, "superspik": [0, 4, 9, 12], "supervis": [0, 10], "learn": [0, 1, 2, 3, 7, 10, 13], "multilay": 0, "neural": [0, 3], "comput": [0, 2, 7, 13], "30": [0, 2, 7], "6": [0, 1, 7, 11], "1514": 0, "1541": 0, "1162": 0, "neco_a_01086": 0, "knight2018": [0, 2, 7], "knight": [0, 3], "j": 0, "nowotni": [0, 3], "gpu": [0, 1, 2, 3, 5, 7, 8, 10], "outperform": 0, "current": [0, 3, 5, 7, 10, 11], "hpc": 0, "neuromorph": 0, "solut": 0, "term": [0, 1, 2, 7], "speed": [0, 1, 7], "energi": 0, "when": [0, 1, 2, 5, 7, 8, 10], "simul": [0, 2, 3, 7, 10, 11], "highli": 0, "connect": [0, 7, 8, 10, 11], "frontier": 0, "neurosci": 0, "12": [0, 7], "decemb": 0, "19": 0, "3389": 0, "fnin": 0, "00941": 0, "turner2022": [0, 2, 7], "turner": 0, "p": [0, 2, 7, 11], "subramanian": 0, "2022": 0, "mlgenn": 0, "acceler": [0, 1, 3], "snn": [0, 2], "infer": 0, "us": [0, 1, 2, 3, 4, 5, 7, 8, 9, 11, 12, 13], "enabl": [0, 1, 5, 11], "engin": 0, "2": [0, 1, 2, 7, 10, 11, 13], "024002": 0, "1088": 0, "2634": 0, "4386": 0, "ac5ac5": 0, "defin": [1, 2, 7, 10], "follow": [1, 2, 5, 7, 10, 11, 13], "gennmodel": [1, 2, 5, 7, 8, 10, 11, 13], "must": [1, 2, 7], "creat": [1, 2, 7, 10, 11, 13], "name": [1, 2, 5, 7], "default": [1, 2, 5, 7, 10, 11, 13], "precis": [1, 2, 7, 11], "see": [1, 2, 7, 11], "ref": 1, "floatprecis": 1, "float": [1, 2, 7, 10, 11, 13], "yourmodelnam": 1, "By": [1, 2, 5, 7], "hardwar": [1, 2, 3, 7, 8], "code": [1, 2, 3, 5, 7, 8, 9, 10, 11, 13], "gener": [1, 3, 5, 7, 9, 10, 11, 13], "backend": [1, 5, 7], "avail": [1, 7], "howev": [1, 2, 5, 7], "thi": [1, 2, 3, 5, 7, 8, 10, 11, 13], "can": [1, 2, 5, 7, 8, 10, 11, 13], "overriden": 1, "keyword": [1, 2, 5, 7], "argument": [1, 2, 5, 7], "For": [1, 2, 5, 7], "exampl": [1, 2, 4, 5, 7, 9, 10, 11, 12, 13], "singl": [1, 2, 5, 7, 8], "thread": [1, 2, 5, 7, 11], "cpu": [1, 2, 5, 7], "could": [1, 2, 5, 7], "manual": [1, 7], "select": [1, 7], "single_threaded_cpu": 1, "run": [1, 2, 5, 7], "smaller": 1, "mai": [1, 2, 7], "fulli": [1, 2, 7], "occupi": 1, "devic": [1, 2, 5, 7, 10, 11], "In": [1, 2, 5, 7, 8, 11], "some": [1, 2, 7], "scenerio": 1, "gradient": [1, 2, 7, 13], "train": [1, 10, 13], "sweep": 1, "overcom": 1, "rune": 1, "multipl": [1, 2, 7], "copi": [1, 2, 5, 7, 10], "same": [1, 2, 5, 7, 8], "batch": [1, 2, 5, 7], "machin": [1, 3], "speak": [1, 7], "genn": [1, 2, 3, 5, 7, 9, 13], "batch_siz": 1, "512": 1, "spars": [1, 5, 7, 8, 10, 11, 13], "ar": [1, 2, 3, 5, 7, 8, 11, 13], "share": [1, 2, 5, 7], "across": [1, 2, 5, 7], "all": [1, 2, 4, 5, 7, 8, 9, 11, 13], "whether": [1, 2, 7], "state": [1, 2, 5, 7, 8, 10], "duplic": [1, 2, 7], "control": 1, "varaccess": [1, 2, 7], "customupdatevaraccess": [1, 2, 7], "associ": [1, 2, 5, 7], "each": [1, 2, 5, 7, 10, 13], "pleas": [1, 2, 3, 7], "todo": [1, 13], "more": [1, 2, 8, 11], "detail": [1, 2, 7], "addition": [1, 2, 5, 7], "ani": [1, 2, 5, 7], "prefer": [1, 7], "expos": [1, 2, 7], "configur": [1, 2, 7], "here": [1, 2, 7], "cuda": [1, 5, 7], "allow": [1, 2, 7, 10, 11], "you": [1, 2, 3, 5, 7, 8], "which": [1, 2, 3, 5, 7], "via": [1, 2, 5, 7], "manual_device_id": [1, 7], "initialis": [1, 7, 8, 13], "constant": [1, 2, 7, 10], "numer": [1, 2, 7], "valu": [1, 2, 7, 10, 11], "homogen": [1, 2], "an": [1, 2, 4, 5, 7, 9, 12], "entir": 1, "ini": 1, "0": [1, 2, 4, 5, 7, 10, 11, 12, 13], "0529324": 1, "thei": [1, 2, 5, 7, 8], "veri": [1, 2, 5, 7], "effici": [1, 2, 7], "access": [1, 2, 5, 7], "from": [1, 2, 3, 4, 5, 7, 10, 11, 12, 13], "either": [1, 7], "hard": 1, "kernel": [1, 2, 5, 7, 11, 13], "deliv": [1, 2, 7, 10], "high": [1, 7], "perform": [1, 2, 7], "cach": [1, 7], "often": [1, 5], "b": [1, 5, 7], "mani": [1, 2, 7, 10, 11], "wai": [1, 2, 7, 8], "fill": 1, "them": [1, 2, 5, 7], "sequenc": [1, 2, 7], "python": [1, 3, 5, 9, 10, 11, 13], "np": [1, 2, 5, 10, 11, 13], "arang": [1, 5, 11], "400": [1, 13], "snippet": [1, 3, 7, 8], "function": [1, 2, 5, 7, 8, 11, 13], "pygenn": [1, 2, 4, 5, 8, 9, 10, 12], "init_var": [1, 7, 11, 13], "param": [1, 2, 7, 10, 13], "union": [1, 2, 7], "initvarsnippetbas": [1, 7], "str": [1, 2, 7, 13], "init": [1, 7, 8, 11, 13], "string": [1, 2, 3, 7, 8], "referenc": [1, 2, 7, 8], "built": [1, 2, 5, 7], "snipept": [1, 7], "init_var_snippet": [1, 6], "instanc": [1, 7], "return": [1, 2, 7, 10, 11, 13], "create_var_init_snippet": [1, 2, 7], "dict": [1, 7], "int": [1, 2, 7, 8, 10, 11, 13], "sampl": [1, 2, 7], "normal": [1, 2, 7, 11], "distribut": [1, 2, 7, 10, 11], "mean": [1, 2, 7, 11, 13], "standard": [1, 2, 7, 10, 11], "deviat": [1, 2, 7, 11], "sd": [1, 2, 7, 11, 13], "dictionari": [1, 5, 11], "As": [1, 2, 5, 7], "well": [1, 2, 5, 7], "variou": [1, 2], "have": [1, 2, 5, 7, 8, 13], "belong": 1, "other": [1, 2, 7, 8], "postsynapt": [1, 5, 7, 8, 11], "attach": [1, 2, 7], "call": [1, 2, 3, 5, 7], "r": [1, 2, 7, 13], "assign": [1, 2, 7], "syntax": 1, "neuron_var_ref": [1, 2, 7], "create_var_ref": [1, 7, 13], "ng": 1, "v": [1, 2, 5, 7, 10, 11, 13], "current_source_var_ref": 1, "cs": [1, 7], "custom_update_var_ref": 1, "cu": 1, "postsynaptic_model_var_ref": 1, "create_psm_var_ref": [1, 7], "sg": [1, 7], "wu_pre_var_ref": 1, "create_wu_pre_var_ref": [1, 7], "pre": [1, 2, 7], "wu_post_var_ref": 1, "create_wu_post_var_ref": [1, 7], "post": [1, 2, 7], "where": [1, 2, 5, 7, 8], "neurongroup": [1, 5, 7], "add_neuron_popul": [1, 7, 10, 11, 13], "currentsourc": [1, 2, 7], "add_current_sourc": [1, 7, 10, 11], "customupd": [1, 7], "add_custom_upd": [1, 7, 13], "synapsegroup": [1, 2, 5, 7, 8], "add_synapse_popul": [1, 7, 8, 10, 11, 13], "while": [1, 2, 5, 7, 11, 13], "interchang": 1, "long": [1, 2, 10], "delai": [1, 2, 5, 7, 8, 11], "size": [1, 2, 7, 10, 13], "per": [1, 2, 5, 7, 11, 13], "weight": [1, 5, 7, 8, 10, 11, 13], "slightli": 1, "differ": [1, 2, 5, 7, 11], "wu_var_ref": 1, "create_wu_var_ref": [1, 7, 13], "g": [1, 2, 5, 7, 10, 11], "cu_wu_var_ref": 1, "customupdatewu": [1, 7], "oper": [1, 2, 5, 7, 8], "anoth": [1, 7], "group": [1, 5, 7, 8], "These": [1, 2, 7], "also": [1, 2, 5, 7], "addit": [1, 2, 7], "featur": [1, 2, 8], "link": 1, "transpos": [1, 7, 13], "wu_transpose_var_ref": 1, "back_sg": 1, "tranpos": [1, 7, 13], "dimens": [1, 7], "i": [1, 2, 7, 8, 10, 11, 13], "e": [1, 2, 5, 7, 11, 13], "its": [1, 7], "_postsynaptic_": 1, "ha": [1, 2, 5, 7], "number": [1, 3, 5, 7, 8, 10, 11, 13], "_presynaptic_": [1, 2, 7], "vice": 1, "versa": 1, "after": [1, 2, 5, 7, 10], "made": [1, 5, 7, 8], "forward": [1, 2, 7], "appli": [1, 2, 5, 7, 11], "_": [1, 2, 5, 7, 11], "onli": [1, 2, 7, 8, 10], "possibl": [1, 2, 7, 8], "synapsematrixtyp": [1, 2, 7, 8], "dens": [1, 7, 8, 10, 13], "onc": [1, 2, 5, 7, 10], "how": [1, 2, 7, 10, 11], "your": [1, 2, 5, 7], "go": [1, 10, 11, 13], "need": [1, 2, 5, 7, 8], "alloc": [1, 5], "memori": [1, 2, 5, 7, 10], "both": [1, 5, 7], "host": [1, 2, 5, 7], "altern": 1, "class": [1, 2, 5, 7], "varloc": [1, 5, 7, 11], "support": [1, 2, 3, 7, 8], "combin": [1, 2, 7], "varlocationattribut": [1, 7], "member": [1, 7], "save": [1, 5, 7, 10, 11, 13], "host_devic": [1, 7], "host_device_zero_copi": [1, 7], "between": [1, 2, 5, 7, 11, 13], "zero": [1, 2, 5, 7, 13], "improv": [1, 7, 8], "data": [1, 2, 5, 7, 10, 11, 13], "frequent": [1, 7], "non": [1, 2, 7], "coher": [1, 7], "architectur": [1, 7], "jetson": [1, 7], "reduc": [1, 2, 7, 8, 13], "note": [1, 3, 7, 11, 13], "newer": 1, "embed": 1, "system": [1, 5], "tx1": 1, "physic": 1, "seper": [1, 7, 8], "thu": [1, 5], "contain": [1, 2, 3, 7], "ad": [1, 2, 7], "pop_nam": [1, 7, 11], "num_neuron": [1, 2, 7, 11], "var": [1, 2, 5, 7, 10, 13], "add": [1, 2, 7, 8, 11, 13], "uniqu": [1, 7, 11], "neuronmodelbas": [1, 7], "neuron_model": [1, 6], "create_neuron_model": [1, 2, 7, 10, 13], "varinit": [1, 7], "ndarrai": [1, 7], "initi": [1, 2, 5, 7, 10, 13], "izhikevich": [1, 7], "set": [1, 2, 5, 7, 8, 10, 11], "tonic": [1, 7], "pop": [1, 5, 7, 10, 11], "02": [1, 2, 7], "65": [1, 7, 11], "d": [1, 7, 11, 13], "u": [1, 7, 11, 13], "20": [1, 7, 10], "two": [1, 2, 7], "Their": 1, "behaviour": [1, 2, 3, 7], "describ": [1, 2, 3, 5, 7, 8], "what": [1, 2, 3, 7], "kind": 1, "dynam": [1, 2, 7, 13], "occur": [1, 2, 7], "output": [1, 2, 7, 10, 11, 13], "presynapt": [1, 2, 7, 13], "typic": [1, 2, 7], "init_weight_upd": [1, 7, 8, 10, 11, 13], "pre_var": [1, 2, 5, 7, 13], "post_var": [1, 2, 5, 7, 13], "pre_var_ref": [1, 2, 7, 13], "post_var_ref": [1, 2, 7, 13], "weight_update_model": [1, 6], "weightupdatemodelbas": [1, 7], "create_weight_update_model": [1, 2, 7, 10, 13], "varrefer": [1, 7], "static": [1, 7], "puls": [1, 7], "weight_init": [1, 7], "staticpulseconstantweight": [1, 7, 8, 10], "input": [1, 2, 7, 10, 11, 13], "translat": [1, 2], "init_postsynapt": [1, 7, 8, 10, 11, 13], "var_ref": [1, 2, 7, 13], "postsynapticmodelbas": [1, 7], "postsynaptic_models_model": [1, 7], "create_postsynaptic_model": [1, 2, 7, 13], "conduct": [1, 7], "exponenti": [1, 2, 7, 13], "shape": [1, 2, 7, 10], "postsynaptic_init": [1, 7], "expcond": [1, 7], "tau": [1, 2, 5, 7, 10, 11, 13], "80": [1, 7, 10, 13], "pop1": [1, 7], "provid": [1, 2, 5, 7], "implement": [1, 2, 4, 7, 8, 9, 12], "dense_proceduralg": [1, 7, 8], "bitmask": [1, 2, 7], "procedur": [1, 2, 7, 8, 11], "procedural_kernelg": [1, 7, 8], "toeplitz": [1, 7, 8], "dense_procedur": 1, "simpli": [1, 7], "init_sparse_connect": [1, 7, 10, 11], "initsparseconnectivitysnippetbas": [1, 7], "init_sparse_connectivity_snippet": [1, 6], "create_sparse_connect_init_snippet": [1, 2, 7], "pair": [1, 2, 7], "probabl": [1, 7, 11], "fixedprob": [1, 7], "prob": [1, 7], "init_toeplitz_connect": [1, 7], "init_toeplitz_connect_snippet": [1, 7], "init_toeplitz_connectivity_snippet": [1, 6], "inittoeplitzconnectivitysnippetbas": [1, 7], "create_toeplitz_connect_init_snippet": [1, 2, 7], "2d": [1, 7], "convolut": [1, 2, 7], "64": [1, 2, 7], "62": [1, 7], "conv_kh": [1, 7], "conv_kw": [1, 7], "conv_ih": [1, 7], "conv_iw": [1, 7], "conv_ic": [1, 7], "conv_oh": [1, 7], "conv_ow": [1, 7], "conv_oc": [1, 7], "conv2d": [1, 7], "should": [1, 2, 7, 10], "4096": [1, 7], "3844": [1, 7], "final": [1, 2, 7], "compon": 1, "place": [1, 5, 7], "matrix_typ": [1, 7, 11], "target": [1, 2, 7, 8, 11, 13], "weight_update_init": [1, 7], "connectivity_init": [1, 7], "none": [1, 2, 7, 10, 11, 13], "sparseconnectivityinit": [1, 7], "toeplitzconnectivityinit": [1, 7], "init_toeplitz_connectivity_connect": [1, 7], "src_pop": [1, 7, 11], "target_pop": [1, 7], "syn": [1, 2, 7], "expcurr": [1, 7, 10, 11, 13], "5": [1, 7, 8, 10, 11, 13], "cs_name": [1, 7], "current_source_model": [1, 6], "currentsourcemodelbas": [1, 7], "create_current_source_model": [1, 2, 7, 10], "inject": [1, 2, 7, 10], "gaussian": [1, 7], "nois": [1, 2, 7, 13], "gaussiannois": [1, 7], "previou": [1, 2, 5, 7, 13], "section": 1, "automat": [1, 2, 7, 11], "everi": [1, 2, 7, 11, 13], "timestep": [1, 2, 5, 7, 10, 11, 13], "process": [1, 5, 8], "would": [1, 2, 5, 7], "benefit": 1, "trigger": [1, 2, 7], "occasion": 1, "classifi": 1, "reset": [1, 2, 7, 10, 13], "stimuli": 1, "been": [1, 7], "present": [1, 10], "optim": [1, 7], "accumul": [1, 2, 7], "over": [1, 2, 5], "sever": [1, 2, 7, 8], "similar": [1, 8], "preced": [1, 7], "calcul": [1, 2, 7, 11, 13], "cu_nam": [1, 7], "group_nam": [1, 7], "custom_update_model": [1, 6], "egp_ref": [1, 7], "includ": [1, 2, 3, 7, 11], "execut": [1, 2, 4, 7, 12], "simultan": [1, 7], "customupdatemodelbas": [1, 7], "create_custom_update_model": [1, 2, 7, 13], "wuvarrefer": [1, 7], "egprefer": [1, 7], "create_egp_ref": [1, 7], "like": [1, 2, 3, 5, 7, 8], "user": [1, 2, 3, 5, 7, 8], "rather": [1, 2, 7, 8, 10, 11, 13], "than": [1, 2, 7, 10, 11, 13], "add_custom_connectivity_upd": [1, 7], "syn_group": [1, 7], "custom_conn_update_model": [1, 7], "customconnectivityupdatemodelbas": [1, 7], "custom_connectivity_update_model": [1, 6], "customconnectivityupdatemodelbaseupdatemodelbas": [1, 7], "create_custom_connectivity_update_model": [1, 2, 7], "One": [2, 5, 7], "main": [2, 7], "thing": [2, 8], "make": [2, 3, 5, 7, 13], "build": [2, 3, 5, 7, 8, 10, 11, 13], "easili": [2, 3, 7], "customis": [2, 3], "languag": [2, 3, 8], "we": [2, 5, 7, 8, 10], "essenti": [2, 7, 13], "c99": [2, 8], "en": 2, "cpprefer": 2, "com": 2, "No": [2, 7], "preprocessor": 2, "enough": 2, "printf": 2, "debug": [2, 7], "messag": 2, "much": [2, 5, 7], "strstr": 2, "etc": [2, 13], "typedefin": 2, "cannot": [2, 3, 5, 7], "esoter": 2, "octal": 2, "integ": [2, 5, 7], "hexadecim": 2, "point": [2, 5, 7, 11, 13], "liter": 2, "aren": 2, "address": [2, 8], "isn": [2, 8], "On": [2, 5], "local": [2, 4, 7, 8, 9, 12], "assum": [2, 8], "store": [2, 7, 8], "regist": [2, 8], "limit": [2, 8], "deal": [2, 8], "extra": [2, 7, 8], "global": [2, 7, 8], "paramet": [2, 7, 8, 10, 11, 13], "arrai": [2, 7, 8, 13], "longer": [2, 7, 8], "do": [2, 5, 7, 8], "stuff": [2, 8], "const": [2, 7, 8, 10, 13], "egpsubset": [2, 8], "egp": [2, 8], "offset": [2, 7, 8, 11], "instead": [2, 7, 8], "overload": [2, 7], "so": [2, 5, 7, 8, 13], "sin": 2, "0f": 2, "resolv": 2, "doubl": 2, "version": [2, 7, 8], "without": 2, "suffix": [2, 7], "treat": [2, 7], "scalar": [2, 7, 10, 13], "alwai": 2, "0d": 2, "lp64": 2, "platform": [2, 5], "32": [2, 7, 13], "bit": 2, "librari": [2, 7], "co": 2, "tan": 2, "aco": 2, "asin": 2, "atan": 2, "atan2": 2, "cosh": 2, "sinh": 2, "tanh": [2, 7], "acosh": 2, "asinh": 2, "atanh": 2, "exp": [2, 7, 10, 13], "expm1": 2, "exp2": 2, "pow": [2, 13], "scalbn": 2, "log": [2, 7, 11, 13], "log1p": 2, "log2": 2, "log10": 2, "ldexp": 2, "ilogb": 2, "sqrt": [2, 11, 13], "cbrt": 2, "hypot": 2, "ceil": [2, 7], "floor": 2, "fmod": 2, "round": [2, 10, 11], "rint": 2, "trunc": 2, "nearbyint": 2, "nextaft": 2, "remaind": 2, "fab": [2, 13], "fdim": 2, "fmax": [2, 7, 10, 13], "fmin": [2, 7, 10, 13], "erf": 2, "erfc": 2, "tgamma": 2, "lgamma": 2, "copysign": 2, "fma": 2, "min": [2, 7, 11, 13], "max": [2, 7, 10, 11, 13], "ab": [2, 11], "form": [2, 7, 8], "probabilist": 2, "mechan": [2, 7], "within": [2, 7, 13], "gennrand_uniform": [2, 7], "drawn": [2, 7], "uniformli": [2, 7], "interv": 2, "gennrand_norm": [2, 7], "gennrand_exponenti": 2, "lambda": [2, 7, 13], "gennrand_log_norm": 2, "std": 2, "specifi": [2, 7], "gennrand_gamma": 2, "alpha": [2, 7, 13], "gamma": [2, 7], "gennrand_binomi": 2, "n": [2, 7, 10, 11], "binomi": [2, 7], "part": [2, 7], "deriv": [2, 5, 7, 13], "popul": [2, 5, 7, 10, 11, 13], "being": [2, 7], "enhanc": 2, "friendli": [2, 5], "decai": [2, 7], "pass": [2, 7, 8], "one": [2, 7], "creation": [2, 11], "bwlo": 2, "derived_param": [2, 7, 13], "exptc": [2, 13], "par": [2, 7, 13], "dt": [2, 7, 10, 11, 13], "new": [2, 7, 8], "class_nam": [2, 7], "var_init_cod": [2, 7], "extra_global_param": [2, 5, 7, 13], "refer": [2, 3, 7, 8], "read": [2, 7], "repres": [2, 7], "step": [2, 5, 7, 10], "And": [2, 7], "id": [2, 7, 11, 13], "index": [2, 3, 7, 11, 13], "start": [2, 7, 13], "synaps": [2, 3, 5, 7, 8, 10, 11, 13], "id_pr": [2, 7], "id_post": [2, 7], "num_pr": [2, 7], "num_post": [2, 7], "write": [2, 7, 8, 11], "option": [2, 7, 8], "tupl": [2, 7], "resolvedtyp": [2, 7], "callabl": [2, 7], "paramss": [2, 7], "statement": [2, 7], "requir": [2, 5, 7, 8, 13], "want": [2, 5, 7, 8, 10], "redraw": [2, 7], "neg": [2, 7], "ensur": [2, 7], "remain": [2, 7], "causal": [2, 7], "normal_positive_model": [2, 7], "normal_posit": [2, 7], "fly": 2, "row_build_cod": [2, 7], "col_build_cod": [2, 7], "calc_max_row_len_func": [2, 7], "calc_max_col_len_func": [2, 7], "calc_kernel_size_func": [2, 7], "id_post_begin": [2, 7], "first": [2, 5, 7, 11, 13], "addsynaps": [2, 7], "x": [2, 5, 7], "row": [2, 7, 8, 13], "column": [2, 7], "maximum": [2, 7, 11], "length": [2, 7], "matrix": [2, 5, 7], "param_nam": [2, 7], "fix": [2, 7, 11, 13], "replac": [2, 7, 8], "scipi": [2, 7, 11], "stat": [2, 7, 11], "import": [2, 5, 7, 10, 11, 13], "binom": [2, 7], "fixed_number_post": [2, 7], "num": [2, 7, 10, 11, 13], "unsign": [2, 7, 13], "idpost": [2, 7], "ppf": [2, 7, 11], "9999": [2, 7, 11], "our": [2, 7], "paper": [2, 7, 11], "short": [2, 7], "end": [2, 7, 10, 11, 13], "up": [2, 3, 7], "text": [2, 7], "frac": [2, 7], "therefor": [2, 7], "look": [2, 3, 5, 7], "invers": [2, 7], "cummul": [2, 7], "cdf": [2, 7], "chanc": [2, 7], "bound": [2, 7], "correct": [2, 7, 10], "draw": [2, 7], "diagonal_build_cod": [2, 7], "diagon": [2, 7, 8], "independ": [2, 7], "id_diag": [2, 7], "id_kern_0": [2, 7], "id_kern_1": [2, 7], "id_kern_n": [2, 7], "dimension": [2, 7], "for_each_synaps": [2, 7, 8], "construct": [2, 3, 5, 7], "loop": [2, 5, 7, 8, 11, 13], "through": [2, 7, 11, 13], "incom": [2, 7], "insid": [2, 7], "convolv": [2, 7], "kern_dim": [2, 7], "squar": [2, 7, 13], "pop_dim": [2, 7], "simple_conv2d_model": [2, 7], "pynn": [2, 7], "simple_conv2d": [2, 7], "kern_siz": [2, 7], "kernrow": [2, 7], "kerncol": [2, 7], "prerow": [2, 7], "precol": [2, 7], "If": [2, 3, 5, 7, 13], "haven": [2, 7], "gone": [2, 7], "off": [2, 7], "edg": [2, 7], "postrow": [2, 7], "postcol": [2, 7], "postind": [2, 7], "express": [2, 7, 13], "extend": [2, 5, 7], "abov": [2, 5, 7, 8], "sim_cod": [2, 7, 10, 13], "threshold_condition_cod": [2, 7, 10, 13], "reset_cod": [2, 7, 10, 13], "additional_input_var": [2, 7, 13], "auto_refractory_requir": [2, 7], "fals": [2, 7, 10, 11, 13], "isyn": [2, 7, 10, 13], "total": [2, 4, 7, 11, 12], "modifi": [2, 7], "threshold": [2, 7], "condit": [2, 7], "test": [2, 7, 10], "emit": [2, 5, 7], "list": [2, 7, 11], "bool": [2, 7], "doe": [2, 7, 8], "auto": [2, 7, 10], "refractori": [2, 7], "logic": [2, 7], "leaki": [2, 7], "integr": [2, 7, 10, 13], "dv": [2, 7], "i_": [2, 7], "rm": [2, 7], "solv": [2, 7], "euler": [2, 7], "method": [2, 5, 7], "leaky_integrator_model": [2, 7], "leaky_integr": [2, 7], "read_writ": [2, 7], "receiv": [2, 7], "linear": [2, 7], "sum": [2, 7, 10, 13], "come": [2, 7], "linearli": [2, 7], "product": [2, 7], "isyn2": [2, 7], "event": [2, 5, 7], "driven": [2, 7], "pre_": [2, 7], "post_": [2, 7], "pre_neuron_var_ref": [2, 7, 13], "post_neuron_var_ref": [2, 7, 13], "pre_spike_syn_cod": [2, 7, 10, 13], "pre_event_syn_cod": [2, 7], "post_event_syn_cod": [2, 7], "post_spike_syn_cod": [2, 7, 10], "synapse_dynamics_cod": [2, 7, 13], "pre_spike_cod": [2, 7, 13], "post_spike_cod": [2, 7], "pre_dynamics_cod": [2, 7, 13], "post_dynamics_cod": [2, 7, 13], "assumpt": [2, 7], "addtopost": [2, 7, 13], "inc": [2, 7], "amount": [2, 7], "dendrit": [2, 7, 11], "insert": [2, 7], "addtopostdelai": [2, 7], "again": [2, 7], "heterogen": [2, 7], "weightupdatemodel": [2, 7], "staticpulsedendriticdelai": [2, 7, 11], "simpl": [2, 3, 7], "max_dendritic_delay_timestep": [2, 7, 11], "properti": [2, 5, 7], "effect": [2, 7], "revers": [2, 7, 11], "direct": [2, 7], "addtopr": [2, 7], "v_post": [2, 7], "_outgoing_": [2, 7], "pre_target_var": [2, 7], "unlik": [2, 7], "action": [2, 7, 10, 11, 13], "modul": [2, 3, 6], "directli": [2, 5, 7, 8], "indic": [2, 5, 7, 11, 13], "varaccessmod": [2, 7], "pre_event_threshold_condition_cod": [2, 7], "post_event_threshold_condition_cod": [2, 7], "stdp": [2, 7, 10], "rule": [2, 7, 10, 13], "nearest": [2, 7], "neighbour": [2, 7], "depend": [2, 7], "delta": [2, 7], "w_": [2, 7], "ij": [2, 7], "begin": [2, 7], "case": [2, 7], "a_": [2, 7], "left": [2, 7], "tau_": [2, 7], "right": [2, 7], "leq0": [2, 7], "manner": [2, 7], "stdp_additive_model": [2, 7], "stdp_addit": [2, 7], "tauplu": [2, 7], "tauminu": [2, 7], "aplu": [2, 7], "aminu": [2, 7], "wmin": [2, 7, 10, 13], "wmax": [2, 7, 10, 13], "st_post": [2, 7, 10], "newweight": [2, 7, 10], "st_pre": [2, 7, 10], "cost": [2, 7], "tend": [2, 7], "grow": [2, 7], "o": [2, 7], "basi": [2, 7], "good": [2, 7], "idea": [2, 7], "pre_var_name_typ": [2, 7], "post_var_name_typ": [2, 7], "_trace_": [2, 7], "stdp_additive_2_model": [2, 7], "genn_model": [2, 7], "create_custom_weight_update_class": [2, 7], "stdp_additive_2": [2, 7], "pretrac": [2, 7], "posttrac": [2, 7], "tauplusdecai": [2, 7], "tauminusdecai": [2, 7], "previous": [2, 5, 7], "intern": [2, 7], "continu": [2, 7], "od": [2, 7], "computation": [2, 7], "costli": [2, 7], "becaus": [2, 5, 7], "larg": [2, 7], "discuss": [2, 5, 7], "context": [2, 7], "rate": [2, 7, 11, 13], "contin": [2, 7], "multipli": [2, 7, 11], "definit": [2, 7], "v_pre": [2, 7], "evalu": [2, 7], "involv": [2, 7], "respect": [2, 7, 8], "voltag": [2, 7], "greater": [2, 7], "whenev": [2, 7], "true": [2, 5, 7, 10, 11, 13], "pre_event_cod": [2, 7], "equat": [2, 7], "injectcurr": [2, 7, 10], "goe": [2, 7], "post_target_var": [2, 7, 13], "injection_cod": [2, 7, 10], "target_var": [2, 7], "uniform_noise_model": [2, 7], "uniform_nois": [2, 7], "magnitud": [2, 7, 10], "demand": 2, "update_cod": [2, 7, 13], "extra_global_param_ref": [2, 7], "launch": [2, 5, 7], "reset_model": [2, 7], "read_onli": [2, 7], "sort": [2, 7, 8, 13], "read_only_dupl": [2, 7], "mode": [2, 7], "reduct": [2, 7], "reduce_batch_sum": [2, 7], "reduce_batch_max": [2, 7], "scenario": [2, 7], "reduce_model": [2, 7], "gradient_batch_reduc": [2, 7], "reducedgradi": [2, 7], "reduce_sum": [2, 7], "reduce_max": [2, 7], "similarli": [2, 5, 7], "reduce_neuron_sum": [2, 7], "reduce_neuron_max": [2, 7], "neuron_reduc": [2, 7], "row_update_cod": [2, 7], "host_update_cod": [2, 7], "design": [2, 7], "issu": [2, 7], "regard": [2, 7], "accompani": [2, 7], "push": [2, 7, 10], "pull": [2, 7], "illustr": [2, 7], "abil": [2, 7], "remov": [2, 7, 8], "remove_diagonal_model": [2, 7], "remove_diagon": [2, 7], "remove_synaps": [2, 7], "break": [2, 7, 13], "back": [2, 7], "add_diagonal_model": [2, 7], "add_diagon": [2, 7], "add_synaps": [2, 7], "lot": [2, 7], "_might_": [2, 7], "detect": [2, 7], "shuffl": [2, 7], "around": [2, 7], "accordingli": [2, 7], "fine": [2, 7], "know": [2, 7], "hook": [2, 7], "abl": [2, 7], "common": [2, 5, 7], "parallel": [2, 7], "determin": [2, 7, 10, 11], "element": [2, 5, 7], "get": [2, 7, 11, 13], "befor": [2, 5, 7], "pushpostindtodevic": [2, 7], "softwar": 3, "packag": [3, 6], "nvidia": 3, "api": [3, 5], "neuron": [3, 5, 7, 10, 11, 13], "genncod": 3, "under": [3, 11, 13], "find": 3, "contact": 3, "project": [3, 10, 11], "develop": [3, 9, 11, 13], "upgrad": 3, "4": [3, 7, 11], "custom": [3, 7, 10, 13], "bibliographi": 3, "maintain": [3, 8], "dr": 3, "jame": 3, "prof": 3, "thoma": 3, "partial": [3, 13], "epsrc": 3, "grant": 3, "ep": 3, "v052241": 3, "unlock": 3, "research": 3, "p006094": 3, "brain": 3, "board": 3, "j019690": 3, "green": 3, "search": 3, "page": 3, "00": [4, 12], "000": [4, 12], "file": [4, 7, 12, 13], "galleri": [4, 9, 10, 11, 13], "mem": [4, 12], "mb": [4, 12], "mnist": [4, 9, 12], "classif": [4, 9, 12], "insect": [4, 9, 12], "inspir": [4, 9, 12], "mushroom": [4, 9, 12], "bodi": [4, 9, 12], "userproject": [4, 7, 12], "mnist_mb_classifi": [4, 10, 12], "py": [4, 10, 11, 12, 13], "potjans_microcircuit": [4, 11, 12], "superspike_demo": [4, 12, 13], "load": [5, 7, 10, 11, 13], "lazi": 5, "hasn": 5, "chang": [5, 7], "almost": [5, 8], "instantan": [5, 7], "error": [5, 7, 13], "report": 5, "simplest": 5, "100": [5, 10, 11], "step_tim": [5, 7, 10, 11, 13], "ms": [5, 7, 11, 13], "asynchron": [5, 11], "synchronis": 5, "natur": [5, 7], "ineffici": [5, 7, 11], "dedic": 5, "collect": 5, "transfer": 5, "chosen": [5, 11], "spike_recording_en": [5, 7, 10, 11, 13], "spike_event_recording_en": [5, 7], "runtim": [5, 6], "num_recording_timestep": [5, 7, 10, 11, 13], "pull_recording_buffers_from_devic": [5, 7, 10, 11, 13], "spike_recording_data": [5, 7, 10, 11, 13], "spike_event_recording_data": 5, "spike_tim": [5, 10, 11], "spike_id": [5, 11], "wa": [5, 8], "real": 5, "interact": 5, "encapsul": 5, "model_preprocessor": [5, 6], "variablebas": [5, 7], "object": [5, 7], "own": 5, "wherea": [5, 8], "locat": [5, 7], "content": 5, "arraybas": [5, 7], "push_to_devic": [5, 7, 10, 13], "pull_from_devic": [5, 7, 10, 13], "noth": 5, "recommend": 5, "leav": 5, "work": [5, 7], "transparantli": 5, "To": [5, 8], "current_valu": [5, 7], "npy": [5, 10], "transform": [5, 7, 13], "format": [5, 7, 13], "matric": 5, "re": [5, 11, 13], "order": [5, 7, 8, 11, 13], "extract": 5, "wish": 5, "current_view": [5, 7], "behav": 5, "extraglobalparamet": [5, 7], "hold": [5, 7], "updat": [5, 7, 8, 11, 13], "psm_extra_global_param": 5, "set_init_valu": [5, 7, 13], "just": [5, 8], "set_param_dynam": [5, 7, 13], "groupmixin": [5, 7], "set_dynamic_param_valu": [5, 7, 13], "increas": [5, 7], "parameterm": 5, "submodul": 6, "cuda_backend": 6, "genn_group": 6, "postsynaptic_model": 6, "single_threaded_cpu_backend": 6, "pybind11_object": 7, "currentsourcemixin": 7, "get_var_loc": 7, "self": 7, "_genn": 7, "arg0": 7, "variabl": [7, 8, 10, 13], "sourc": [7, 9, 10, 11, 13], "set_var_loc": 7, "arg1": 7, "var_initialis": 7, "customconnectivityupd": 7, "customconnectivityupdatemixin": 7, "get_post_var_loc": 7, "get_pre_var_loc": 7, "post_var_initialis": 7, "post_var_refer": 7, "pre_var_initialis": 7, "pre_var_refer": 7, "set_post_var_loc": 7, "ignor": 7, "space": 7, "set_pre_var_loc": 7, "synapse_group": 7, "update_group_nam": 7, "var_refer": 7, "customupdatebas": 7, "customupdatemixin": 7, "ax": [7, 11, 13], "subtract": 7, "ie": 7, "varaccessdim": 7, "axi": [7, 10, 13], "written": 7, "whatev": [7, 8], "read_only_shar": 7, "asid": 7, "read_only_shared_neuron": 7, "33": 7, "84": 7, "76": 7, "52": 7, "44": 7, "customupdatewumixin": 7, "model_nam": 7, "time_precis": 7, "genn_log_level": 7, "plogsever": 7, "warn": 7, "code_gen_log_level": 7, "transpiler_log_level": 7, "runtime_log_level": 7, "backend_log_level": 7, "preference_kwarg": 7, "modelspecintern": 7, "interfac": 7, "backend_nam": 7, "path_to_model": 7, "always_rebuild": 7, "never_rebuild": 7, "path": 7, "directori": 7, "rebuilt": 7, "even": 7, "doesn": [7, 11], "appear": 7, "never": 7, "ever": 7, "prevent": 7, "overwrit": 7, "custom_upd": [7, 13], "get_custom_update_tim": [7, 13], "second": [7, 11, 13], "spent": 7, "kernel_tim": 7, "get_custom_update_transpose_tim": [7, 13], "init_sparse_tim": [7, 11, 13], "init_tim": [7, 11, 13], "record": [7, 10, 11, 13], "neuron_update_tim": [7, 11, 13], "postsynaptic_update_tim": 7, "presynaptic_update_tim": [7, 11, 13], "buffer": 7, "synapse_dynamics_tim": [7, 13], "unload": 7, "free": 7, "neurongroupmixin": 7, "num_delay_slot": 7, "prev_spike_time_loc": 7, "prev_spike_time_requir": 7, "recording_zero_copy_en": 7, "spike_time_loc": 7, "spike_time_requir": 7, "parallelismhint": 7, "word_packed_bitmask": 7, "fatal": 7, "info": 7, "verbos": 7, "synapsegroupmixin": 7, "axonal_delay_step": [7, 8], "back_prop_delay_step": 7, "dendritic_delay_loc": 7, "get_ps_var_loc": 7, "get_wu_post_var_loc": 7, "get_wu_pre_var_loc": 7, "get_wu_var_loc": 7, "kernel_s": 7, "max_connect": 7, "max_source_connect": 7, "narrow_sparse_ind_en": 7, "num_threads_per_spik": [7, 11], "output_loc": 7, "parallelism_hint": 7, "ps_initialis": 7, "set_ps_param_dynam": 7, "set_ps_var_loc": 7, "set_wu_param_dynam": 7, "set_wu_post_var_loc": 7, "set_wu_pre_var_loc": 7, "set_wu_var_loc": 7, "sparse_connectivity_initialis": 7, "sparse_connectivity_loc": 7, "toeplitz_connectivity_initialis": 7, "wu_initialis": 7, "synapsematrixconnect": 7, "flag": 7, "differnet": 7, "8": [7, 11], "16": 7, "66": 7, "129": 7, "136": 7, "264": 7, "68": 7, "272": 7, "synapsematrixweight": 7, "individu": 7, "256": [7, 13], "128": 7, "97": 7, "varaccessmodeattribut": 7, "attribut": 7, "summat": 7, "7": [7, 11], "zero_copi": 7, "arg": [7, 10, 11, 13], "kwarg": 7, "create_psm_egp_ref": 7, "create_wu_egp_ref": 7, "var_nam": [7, 10], "transpose_sg": 7, "transpose_var_nam": 7, "get_var_access_dim": 7, "blocksizeselect": 7, "block": 7, "occup": 7, "pick": 7, "blocksiz": 7, "deviceselect": 7, "most_memori": 7, "most": [7, 8], "preferencesbas": 7, "block_size_select_method": 7, "constant_cache_overhead": 7, "alreadi": 7, "four": 7, "header": [7, 11, 13], "neuronupd": 7, "synapseupd": 7, "runner": 7, "take": [7, 8], "72": 7, "byte": 7, "lookup": 7, "tabl": 7, "curand": 7, "applic": 7, "device_select_method": 7, "enable_nccl_reduct": 7, "correspond": 7, "nccl": 7, "generate_line_info": 7, "line": 7, "result": 7, "profil": [7, 11, 13], "purpos": 7, "manual_block_s": 7, "show_ptx_info": 7, "ptx": 7, "assembl": 7, "inform": 7, "displai": [7, 13], "dure": 7, "compil": [7, 8], "dc": [7, 11], "It": [7, 11, 13], "amp": 7, "amplitud": 7, "na": [7, 10, 11], "noisi": 7, "poissonexp": [7, 11], "equival": 7, "poisson": [7, 13], "tausyn": [7, 11], "fire": [7, 10, 11], "hz": [7, 11], "minim": [7, 10], "get_var_valu": 7, "parent": 7, "pull_extra_global_param_from_devic": 7, "egp_nam": 7, "pull_var_from_devic": 7, "given": [7, 11], "push_extra_global_param_to_devic": 7, "push_var_to_devic": 7, "get_sparse_post_ind": [7, 10], "get_sparse_pre_ind": [7, 10], "post_spike_event_recording_data": 7, "pre_spike_event_recording_data": 7, "pull_connectivity_from_devic": [7, 10], "pull_in_syn_from_devic": 7, "pull_psm_extra_global_param_from_devic": 7, "wrapper": 7, "push_connectivity_to_devic": 7, "push_in_syn_to_devic": 7, "push_psm_extra_global_param_to_devic": 7, "set_sparse_connect": [7, 10], "pre_indic": 7, "post_indic": 7, "rag": 7, "weight_update_var_s": 7, "convert": [7, 10, 13], "channel": 7, "rang": [7, 10, 13], "fixednumberpostwithreplac": 7, "random": [7, 13], "discret": 7, "uniform": 7, "ascend": 7, "1st": 7, "statist": 7, "beta": [7, 13], "npost": 7, "next": 7, "smallest": 7, "special": 7, "fixednumberprewithreplac": [7, 10], "fixednumbertotalwithreplac": [7, 11], "stage": 7, "multinomi": 7, "parallelis": 7, "exist": 7, "bernoulli": 7, "repeatedli": 7, "geometr": 7, "trial": [7, 13], "success": 7, "gap": 7, "devroy": 7, "1986": 7, "invert": 7, "fixedprobabilitynoautaps": 7, "autaps": 7, "recurr": 7, "br": 7, "inneffici": 7, "gemetr": 7, "onetoon": 7, "uninitialis": 7, "mark": 7, "avgpoolconv2d": 7, "averag": [7, 10, 13], "pool": 7, "intialis": 7, "seldom": 7, "initvarsnippet": 7, "implicit": [7, 8], "constructor": 7, "unit": [7, 13], "distanc": 7, "normalclip": [7, 11, 13], "resampl": 7, "out": 7, "my": 7, "thgenn": 7, "minimum": 7, "normalclippeddelai": [7, 11], "intend": 7, "variable_typ": 7, "view": [7, 10, 13], "set_arrai": 7, "view_shap": 7, "variable_nam": 7, "init_valu": 7, "about": 7, "iter": 7, "set_valu": 7, "synapsevari": 7, "delay_group": 7, "valinit_valuesu": 7, "cite": 7, "izhikevich2003simpl": 7, "usual": 7, "eqnarrai": 7, "04": 7, "140": 7, "du": 7, "bv": 7, "extern": [7, 11], "increment": 7, "mv": 7, "particular": 7, "popular": 7, "though": 7, "due": 7, "strictli": 7, "inconsist": 7, "membran": [7, 13], "potenti": [7, 13], "recoveri": 7, "sensit": 7, "izhikevichvari": 7, "neuronmodel": 7, "lif": [7, 10, 11], "vrest": [7, 10, 11, 13], "unless": 7, "randomli": 7, "vspike": 7, "spiketim": [7, 13], "trefract": 7, "last": 7, "period": 7, "tspike": 7, "durat": [7, 11], "rest": 7, "entri": [7, 11, 13], "That": 7, "equal": 7, "undefin": 7, "firingprob": 7, "cdot": 7, "pattern": 7, "leq": 7, "approxim": 7, "math": 7, "relev": 7, "especi": 7, "quit": 7, "small": 7, "moder": 7, "worth": 7, "becom": [7, 8], "poor": 7, "poissonnew": 7, "accord": 7, "timesteptospik": 7, "11": 7, "rulkovmap": 7, "rulkov": 7, "map": 7, "rulkov2002": 7, "nowotny2005self": 7, "ll": 7, "v_": 7, "big": 7, "y": 7, "otherwis": [7, 13], "prev": 7, "60mv": 7, "shift": 7, "excit": 7, "origin": [7, 11], "468": 7, "roughli": 7, "resist": 7, "regul": 7, "omega": 7, "spikesourc": 7, "empti": 7, "spikegeneratorgroup": 7, "brian": 7, "spikesourcearrai": [7, 13], "startspik": [7, 13], "endspik": [7, 13], "globel": 7, "traubmil": 7, "hodgkin": 7, "huxlei": 7, "traub": 7, "mile": 7, "algorithm": 7, "taken": 7, "traub1991": 7, "i_k": 7, "leak": 7, "i_m": 7, "i_i": 7, "g_": 7, "m_i": 7, "h_i": 7, "v_i": 7, "e_": 7, "k": 7, "n_i": 7, "dy": 7, "alpha_i": 7, "beta_i": 7, "y_i": 7, "h": [7, 10, 11, 13], "alpha_n": 7, "032": 7, "50": [7, 10, 11, 13], "beta_n": 7, "55": 7, "40": 7, "alpha_m": 7, "beta_m": 7, "28": 7, "25": [7, 11], "alpha_h": 7, "48": 7, "18": 7, "beta_h": 7, "143": 7, "nf": 7, "02672": 7, "mu": 7, "63": 7, "563": 7, "15": [7, 10], "43": 7, "95": 7, "gna": 7, "mohm": 7, "cm": 7, "ena": 7, "equi": 7, "gk": 7, "ek": 7, "gl": 7, "el": 7, "capac": [7, 10], "densiti": 7, "muf": 7, "ordinari": 7, "differenti": 7, "ldt": 7, "004": 7, "variant": [7, 8], "IF": [7, 10], "check": 7, "singular": 7, "hit": 7, "l": 7, "hospit": 7, "traubmilesalt": 7, "workaround": 7, "avoid": 7, "munimum": 7, "traubmilesfast": 7, "fast": 7, "inner": 7, "There": 7, "traubmilesnstep": 7, "deltacurr": [7, 10, 13], "Is": 7, "expdecai": 7, "expf": 7, "piecewisestdp": 7, "finit": 7, "transmiss": 7, "window": 7, "piecewis": 7, "imag": [7, 10], "html": 7, "learn1synapse_explain_html": 7, "png": 7, "latex": 7, "learn1synapse_explain": 7, "width": 7, "10cm": 7, "curv": 7, "raw": 7, "graw": 7, "filter": [7, 13], "sugmoid": 7, "obtain": 7, "impli": 7, "unpair": 7, "incur": 7, "henc": 7, "stxx": 7, "xx": [7, 8], "somewhat": [7, 11], "arbitrarili": 7, "subject": 7, "sigmoid": 7, "revert": 7, "correctli": 7, "map_classol": 7, "cc": 7, "mbody1": 7, "neuronn": 7, "gkcdn": 7, "scalar_min": 7, "cnt": 7, "fprintf": 7, "stdout": 7, "too": 7, "low": [7, 11], "below": [7, 9], "tmp": 7, "mykcdn_p": 7, "grawkcdn": 7, "cerr": 7, "endl": 7, "lead": 7, "infin": 7, "nomin": 7, "extrem": 7, "act": 7, "g_0": 7, "t_": 7, "compar": 7, "figur": [7, 11], "tlrn": 7, "tchng": 7, "tdecai": 7, "strength": 7, "tpunish10": 7, "suppress": 7, "respons": 7, "tpunish01": 7, "gmax": 7, "maxim": 7, "achiev": [7, 8], "gmid": 7, "midpoint": 7, "gslope": 7, "slope": 7, "taushift": 7, "gsyn0": 7, "staticgrad": 7, "grade": 7, "gradual": 7, "gsyn": 7, "larger": 7, "epr": 7, "vslope": 7, "releas": 7, "_pre": 7, "_post": 7, "staticpuls": [7, 8, 10], "coupl": 7, "sim": 7, "aim": 8, "backward": 8, "compat": 8, "strive": 8, "underli": 8, "pars": 8, "subset": 8, "old": 8, "necessari": [8, 11], "streamlin": 8, "area": 8, "psm": 8, "cumbersom": 8, "wors": 8, "now": 8, "axon": 8, "were": 8, "realli": 8, "ugli": 8, "confus": 8, "let": 8, "outsid": 8, "reus": 8, "globalg": 8, "individualg": 8, "Then": 8, "ve": 8, "renam": 8, "obviou": 8, "chose": 8, "unusu": 8, "pointer": 8, "await": 8, "pr": 8, "settabl": 8, "refernec": 8, "download": [9, 10, 11, 13], "userproject_python": 9, "zip": [9, 13], "jupyt": [9, 10, 11, 13], "notebook": [9, 10, 11, 13], "userproject_jupyt": 9, "sphinx": [9, 10, 11, 13], "digit": 10, "usag": [10, 11, 13], "plot": [10, 11, 13], "numpi": [10, 11, 13], "argpars": [10, 11, 13], "argumentpars": [10, 11, 13], "tqdm": 10, "factor": [10, 11], "normalis": 10, "pixel": 10, "input_scal": 10, "stimul": 10, "mbon": 10, "mbon_stimulus_curr": 10, "match": [10, 11], "num_pn": 10, "784": 10, "kenyon": 10, "num_kc": 10, "20000": 10, "num_mbon": 10, "present_time_m": 10, "lif_param": [10, 11], "taum": [10, 11], "60": [10, 13], "vreset": [10, 11], "vthresh": [10, 11, 13], "ioffset": [10, 11], "taurefrac": [10, 11, 13], "pn": 10, "pn_param": 10, "pn_kc_weight": 10, "pn_kc_tau_syn": 10, "pn_kc_fan_in": 10, "kc": 10, "ggn": 10, "inhibit": 10, "200": [10, 13], "ggn_param": 10, "kc_mbon_tau_syn": 10, "kc_mbon_param": 10, "rho": 10, "01": 10, "eta": 10, "00002": 10, "cs_model": 10, "if_model": 10, "symmetr": 10, "symmetric_stdp": 10, "cli": [10, 13], "def": [10, 11, 13], "get_pars": [10, 11, 13], "parser": [10, 11, 13], "add_argu": [10, 11, 13], "store_tru": [10, 11, 13], "help": [10, 11, 13], "__name__": [10, 11, 13], "__main__": [10, 11, 13], "parse_arg": [10, 11, 13], "reshap": [10, 13], "test_imag": 10, "els": [10, 11, 13], "train_imag": 10, "astyp": 10, "float32": [10, 11], "newaxi": 10, "label": 10, "test_label": 10, "train_label": 10, "mnist_mb": 10, "lif_init": [10, 11], "refractim": [10, 11, 13], "if_init": 10, "turn": [10, 13], "pn_input": 10, "mbon_input": 10, "pn_kc_connect": 10, "pn_kc": 10, "pn_kc_ind": 10, "kc_ggn": 10, "ggn_kc": 10, "kc_mbon_weight_upd": 10, "kc_mbon_g": 10, "kc_mbon": 10, "present_timestep": 10, "reset_spike_tim": 10, "finfo": [10, 11], "reset_out_post": 10, "out_post": 10, "reset_neuron": 10, "var_init": 10, "var_val": 10, "item": [10, 11], "num_correct": 10, "count": [10, 13], "mbon_spike_tim": 10, "mbon_spike_id": 10, "len": [10, 11, 13], "argmin": 10, "print": [10, 11, 13], "weigh": 10, "kc_mbon_g_view": 10, "vstack": [10, 13], "plot_weight_distribut": 10, "matplotlib": [10, 11, 13], "pyplot": [10, 11, 13], "plt": [10, 11, 13], "fig": [10, 11, 13], "subplot": [10, 11, 13], "figsiz": 10, "hist": 10, "bin": 10, "axvlin": 10, "linestyl": 10, "set_xlabel": [10, 11, 13], "set_ylabel": [10, 11, 13], "show": [10, 11, 13], "ipynb": [10, 11, 13], "reimplement": [11, 13], "tobia": 11, "marku": 11, "spontan": 11, "irregular": 11, "agreement": 11, "vivo": 11, "awak": 11, "anim": 11, "layer": 11, "excitatori": 11, "neuron_scal": 11, "connectivity_scal": 11, "1000": [11, 13], "norm": 11, "perf_count": 11, "layer_nam": 11, "23": 11, "population_nam": 11, "dt_m": 11, "background": 11, "background_r": 11, "rel": 11, "inhibitori": 11, "except": 11, "l4e": 11, "l2": 11, "3e": 11, "mean_w": 11, "87": 11, "8e": 11, "external_w": 11, "801": 11, "paragraph": 11, "parameter": 11, "caption": 11, "supplementari": 11, "layer_23_4_w": 11, "rel_w": 11, "mention": 11, "layer_23_4_relw": 11, "05": [11, 13], "20683": 11, "5834": 11, "21915": 11, "5479": 11, "4850": 11, "1065": 11, "14395": 11, "2948": 11, "connection_probabilti": 11, "23e": 11, "1009": 11, "23i": 11, "1689": 11, "4e": 11, "0437": 11, "4i": 11, "0818": 11, "5e": 11, "0323": 11, "5i": 11, "6e": 11, "0076": 11, "6i": 11, "1346": 11, "1371": 11, "0316": 11, "0515": 11, "0755": 11, "0042": 11, "0077": 11, "0059": 11, "0497": 11, "135": 11, "0067": 11, "0003": 11, "0453": 11, "0691": 11, "0029": 11, "0794": 11, "1597": 11, "0033": 11, "1057": 11, "1004": 11, "0622": 11, "0505": 11, "0057": 11, "0831": 11, "3726": 11, "0204": 11, "0548": 11, "0269": 11, "0257": 11, "0022": 11, "06": 11, "3158": 11, "0086": 11, "0156": 11, "0066": 11, "0211": 11, "0166": 11, "0572": 11, "0197": 11, "0396": 11, "2252": 11, "0364": 11, "001": [11, 13], "0034": 11, "0005": 11, "0277": 11, "0658": 11, "1443": 11, "degre": 11, "num_external_input": 11, "1600": 11, "1500": 11, "2100": 11, "1900": 11, "2000": 11, "2900": 11, "realiz": 11, "mean_firing_r": 11, "971": 11, "868": 11, "746": 11, "396": 11, "142": 11, "9": [11, 13], "078": 11, "991": 11, "523": 11, "mean_delai": 11, "75": 11, "delay_sd": 11, "375": 11, "helper": [11, 13], "get_scaled_num_neuron": 11, "get_full_num_input": 11, "src_layer": 11, "trg_layer": 11, "trg_pop": 11, "num_src": 11, "num_trg": 11, "connection_prob": 11, "get_mean_weight": 11, "get_scaled_num_connect": 11, "num_input": [11, 13], "assert": 11, "get_full_mean_input_curr": 11, "mean_input_curr": 11, "fuse_postsynaptic_model": 11, "default_narrow_sparse_ind_en": 11, "timing_en": [11, 13], "kernel_profil": [11, 13], "default_var_loc": 11, "default_sparse_connectivity_loc": 11, "58": 11, "poisson_init": 11, "exp_curr_init": 11, "quantil": 11, "normal_quantile_cdf": 11, "max_delai": 11, "fm": 11, "slot": 11, "seem": 11, "aggress": 11, "merg": 11, "max_dendritic_delay_slot": 11, "total_neuron": 11, "neuron_popul": 11, "ext_input_r": 11, "ext_weight": 11, "ext_input_curr": 11, "poisson_param": 11, "pop_siz": 11, "neuron_pop": 11, "_poisson": 11, "tpopul": 11, "total_synaps": 11, "num_sub_row": 11, "procedural_connect": 11, "trg_name": 11, "src_name": 11, "mean_weight": 11, "weight_sd": 11, "num_connect": 11, "tconnect": 11, "numconnect": 11, "meanweight": 11, "weightsd": 11, "meandelai": 11, "delaysd": 11, "connector": 11, "connect_param": 11, "d_dist": 11, "synapse_nam": 11, "hack": 11, "cast": 11, "w_dist": 11, "static_synapse_init": 11, "syn_pop": 11, "span": 11, "duration_timestep": 11, "ten_percent_timestep": 11, "sim_start_tim": 11, "advanc": 11, "sim_end_tim": 11, "tsimul": 11, "tinit": 11, "tspars": 11, "tneuron": 11, "tsynaps": 11, "save_data": [11, 13], "csv": [11, 13], "savetxt": [11, 13], "_spike": 11, "column_stack": [11, 13], "delimit": [11, 13], "fmt": [11, 13], "yuck": 11, "ordered_neuron_popul": 11, "start_id": 11, "bar_i": 11, "actor": 11, "scatter": [11, 13], "edgecolor": [11, 13], "bar": 11, "colour": 11, "barh": 11, "align": 11, "center": 11, "color": 11, "get_facecolor": 11, "ecolor": 11, "black": 11, "po": 11, "firingr": 11, "set_ytick": 11, "set_yticklabel": 11, "friedemann": 13, "surya": 13, "radcliff": 13, "camera": 13, "oxford": 13, "record_tri": 13, "target_fil": 13, "num_trial": 13, "filenam": 13, "ra": 13, "600": 13, "timestep_m": 13, "num_output": 13, "num_hidden": 13, "tau_rise_m": 13, "tau_decay_m": 13, "tau_rms_m": 13, "30000": 13, "tau_avg_err_m": 13, "10000": 13, "r0": 13, "epsilon": 13, "1e": 13, "tau_decay_": 13, "tau_rise_": 13, "tau_avg_err_": 13, "scale_tr_err_flt": 13, "auryn": 13, "volt": 13, "1000x": 13, "w_min": 13, "w_max": 13, "w0": 13, "experi": 13, "input_freq_hz": 13, "update_time_m": 13, "500": 13, "trial_m": 13, "1890": 13, "update_timestep": 13, "trial_timestep": 13, "calc_t_peak": 13, "tau_ris": 13, "tau_decai": 13, "write_spike_fil": 13, "r_max_prop_model": 13, "r_max_prop": 13, "updatetim": 13, "taurm": 13, "upsilon": 13, "updatetimestep": 13, "exprm": 13, "superspike_model": 13, "tauris": 13, "taudecai": 13, "z": 13, "ztilda": 13, "sigmaprim": 13, "errtilda": 13, "trace": 13, "oneplushi": 13, "elig": 13, "feedback_model": 13, "feedback": 13, "hidden_neuron_model": 13, "hidden": 13, "taumem": 13, "isynfeedback": 13, "rmembran": 13, "output_neuron_model": 13, "tauavgerr": 13, "errris": 13, "avgsqrerr": 13, "errdecai": 13, "normfactor": 13, "trisemult": 13, "tdecaymult": 13, "tpeak": 13, "mulavgerr": 13, "spred": 13, "sreal": 13, "mismatch": 13, "temp": 13, "narg": 13, "target_spik": 13, "loadtxt": 13, "dtype": 13, "neuron_id": 13, "millisecond": 13, "target_neuron_end_tim": 13, "cumsum": 13, "bincount": 13, "minlength": 13, "target_neuron_start_tim": 13, "concaten": 13, "frozen": 13, "input_isi_m": 13, "input_spike_tim": 13, "vector": 13, "reach": 13, "stack": 13, "input_spikes_per_neuron": 13, "togeth": 13, "input_spik": 13, "input_neuron_end_tim": 13, "input_neuron_start_tim": 13, "input_init_var": 13, "hidden_param": 13, "hidden_init_var": 13, "output_param": 13, "output_init_var": 13, "superspike_param": 13, "superspike_pre_init_var": 13, "superspike_post_init_var": 13, "input_hidden_weight_dist_param": 13, "input_hidden_init_var": 13, "hidden_output_weight_dist_param": 13, "hidden_output_init_var": 13, "r_max_prop_param": 13, "descript": 13, "generatelineinfo": 13, "any_record": 13, "input_hidden": 13, "inputhidden": 13, "hidden_output": 13, "hiddenoutput": 13, "output_hidden": 13, "outputhidden": 13, "input_hidden_transpos": 13, "calculatetranspos": 13, "input_hidden_optimiser_var_ref": 13, "input_hidden_optimis": 13, "gradientlearn": 13, "hidden_output_optimiser_var_ref": 13, "hidden_output_optimis": 13, "output_avg_sqr_err_var": 13, "current_r0": 13, "hidden_spik": 13, "output_spik": 13, "perid": 13, "time_": 13, "mean_error": 13, "0e": 13, "upload": 13, "repeat": 13, "input_spikes_": 13, "hidden_spikes_": 13, "output_spikes_": 13, "append": 13, "sharex": 13, "col": 13, "sharei": 13, "enumer": 13, "start_time_": 13, "890": 13}, "objects": {"": [[7, 0, 0, "-", "pygenn"]], "pygenn": [[7, 1, 1, "", "CurrentSource"], [7, 1, 1, "", "CustomConnectivityUpdate"], [7, 1, 1, "", "CustomUpdate"], [7, 1, 1, "", "CustomUpdateVarAccess"], [7, 1, 1, "", "CustomUpdateWU"], [7, 1, 1, "", "GeNNModel"], [7, 1, 1, "", "NeuronGroup"], [7, 1, 1, "", "ParallelismHint"], [7, 1, 1, "", "PlogSeverity"], [7, 1, 1, "", "SynapseGroup"], [7, 1, 1, "", "SynapseMatrixConnectivity"], [7, 1, 1, "", "SynapseMatrixType"], [7, 1, 1, "", "SynapseMatrixWeight"], [7, 1, 1, "", "VarAccess"], [7, 1, 1, "", "VarAccessDim"], [7, 1, 1, "", "VarAccessMode"], [7, 1, 1, "", "VarAccessModeAttribute"], [7, 1, 1, "", "VarLocation"], [7, 1, 1, "", "VarLocationAttribute"], [7, 5, 1, "", "create_current_source_model"], [7, 5, 1, "", "create_custom_connectivity_update_model"], [7, 5, 1, "", "create_custom_update_model"], [7, 5, 1, "", "create_egp_ref"], [7, 5, 1, "", "create_neuron_model"], [7, 5, 1, "", "create_postsynaptic_model"], [7, 5, 1, "", "create_psm_egp_ref"], [7, 5, 1, "", "create_psm_var_ref"], [7, 5, 1, "", "create_sparse_connect_init_snippet"], [7, 5, 1, "", "create_toeplitz_connect_init_snippet"], [7, 5, 1, "", "create_var_init_snippet"], [7, 5, 1, "", "create_var_ref"], [7, 5, 1, "", "create_weight_update_model"], [7, 5, 1, "", "create_wu_egp_ref"], [7, 5, 1, "", "create_wu_post_var_ref"], [7, 5, 1, "", "create_wu_pre_var_ref"], [7, 5, 1, "", "create_wu_var_ref"], [7, 0, 0, "-", "cuda_backend"], [7, 0, 0, "-", "current_source_models"], [7, 0, 0, "-", "custom_connectivity_update_models"], [7, 0, 0, "-", "custom_update_models"], [7, 0, 0, "-", "genn_groups"], [7, 5, 1, "", "get_var_access_dim"], [7, 5, 1, "", "init_postsynaptic"], [7, 5, 1, "", "init_sparse_connectivity"], [7, 0, 0, "-", "init_sparse_connectivity_snippets"], [7, 5, 1, "", "init_toeplitz_connectivity"], [7, 0, 0, "-", "init_toeplitz_connectivity_snippets"], [7, 5, 1, "", "init_var"], [7, 0, 0, "-", "init_var_snippets"], [7, 5, 1, "", "init_weight_update"], [7, 0, 0, "-", "model_preprocessor"], [7, 0, 0, "-", "neuron_models"], [7, 0, 0, "-", "postsynaptic_models"], [7, 0, 0, "-", "single_threaded_cpu_backend"], [7, 0, 0, "-", "types"], [7, 0, 0, "-", "weight_update_models"]], "pygenn.CurrentSource": [[7, 2, 1, "", "get_var_location"], [7, 3, 1, "", "model"], [7, 3, 1, "", "name"], [7, 3, 1, "", "params"], [7, 2, 1, "", "set_param_dynamic"], [7, 2, 1, "", "set_var_location"], [7, 3, 1, "", "var_initialisers"]], "pygenn.CustomConnectivityUpdate": [[7, 2, 1, "", "get_post_var_location"], [7, 2, 1, "", "get_pre_var_location"], [7, 2, 1, "", "get_var_location"], [7, 3, 1, "", "model"], [7, 3, 1, "", "name"], [7, 3, 1, "", "params"], [7, 3, 1, "", "post_var_initialisers"], [7, 3, 1, "", "post_var_references"], [7, 3, 1, "", "pre_var_initialisers"], [7, 3, 1, "", "pre_var_references"], [7, 2, 1, "", "set_param_dynamic"], [7, 2, 1, "", "set_post_var_location"], [7, 2, 1, "", "set_pre_var_location"], [7, 2, 1, "", "set_var_location"], [7, 3, 1, "", "synapse_group"], [7, 3, 1, "", "update_group_name"], [7, 3, 1, "", "var_initialisers"], [7, 3, 1, "", "var_references"]], "pygenn.CustomUpdate": [[7, 3, 1, "", "num_neurons"], [7, 3, 1, "", "var_references"]], "pygenn.CustomUpdateVarAccess": [[7, 4, 1, "", "READ_ONLY"], [7, 4, 1, "", "READ_ONLY_SHARED"], [7, 4, 1, "", "READ_ONLY_SHARED_NEURON"], [7, 4, 1, "", "READ_WRITE"], [7, 4, 1, "", "REDUCE_BATCH_MAX"], [7, 4, 1, "", "REDUCE_BATCH_SUM"], [7, 4, 1, "", "REDUCE_NEURON_MAX"], [7, 4, 1, "", "REDUCE_NEURON_SUM"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.CustomUpdateWU": [[7, 3, 1, "", "synapse_group"], [7, 3, 1, "", "var_references"]], "pygenn.GeNNModel": [[7, 2, 1, "", "add_current_source"], [7, 2, 1, "", "add_custom_connectivity_update"], [7, 2, 1, "", "add_custom_update"], [7, 2, 1, "", "add_neuron_population"], [7, 2, 1, "", "add_synapse_population"], [7, 3, 1, "", "backend_name"], [7, 2, 1, "", "build"], [7, 2, 1, "", "custom_update"], [7, 3, 1, "", "dT"], [7, 2, 1, "", "get_custom_update_time"], [7, 2, 1, "", "get_custom_update_transpose_time"], [7, 3, 1, "", "init_sparse_time"], [7, 3, 1, "", "init_time"], [7, 2, 1, "", "load"], [7, 3, 1, "", "neuron_update_time"], [7, 3, 1, "", "postsynaptic_update_time"], [7, 3, 1, "", "presynaptic_update_time"], [7, 2, 1, "", "pull_recording_buffers_from_device"], [7, 2, 1, "", "step_time"], [7, 3, 1, "", "synapse_dynamics_time"], [7, 3, 1, "", "t"], [7, 3, 1, "", "timestep"], [7, 2, 1, "", "unload"]], "pygenn.NeuronGroup": [[7, 2, 1, "", "get_var_location"], [7, 3, 1, "", "model"], [7, 3, 1, "", "name"], [7, 3, 1, "", "num_delay_slots"], [7, 3, 1, "", "num_neurons"], [7, 3, 1, "", "params"], [7, 3, 1, "", "prev_spike_time_location"], [7, 3, 1, "", "prev_spike_time_required"], [7, 3, 1, "", "recording_zero_copy_enabled"], [7, 2, 1, "", "set_param_dynamic"], [7, 2, 1, "", "set_var_location"], [7, 3, 1, "", "spike_event_recording_enabled"], [7, 3, 1, "", "spike_recording_enabled"], [7, 3, 1, "", "spike_time_location"], [7, 3, 1, "", "spike_time_required"], [7, 3, 1, "", "var_initialisers"]], "pygenn.ParallelismHint": [[7, 4, 1, "", "POSTSYNAPTIC"], [7, 4, 1, "", "PRESYNAPTIC"], [7, 4, 1, "", "WORD_PACKED_BITMASK"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.PlogSeverity": [[7, 4, 1, "", "DEBUG"], [7, 4, 1, "", "ERROR"], [7, 4, 1, "", "FATAL"], [7, 4, 1, "", "INFO"], [7, 4, 1, "", "NONE"], [7, 4, 1, "", "VERBOSE"], [7, 4, 1, "", "WARNING"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.SynapseGroup": [[7, 3, 1, "", "axonal_delay_steps"], [7, 3, 1, "", "back_prop_delay_steps"], [7, 3, 1, "", "dendritic_delay_location"], [7, 2, 1, "", "get_ps_var_location"], [7, 2, 1, "", "get_wu_post_var_location"], [7, 2, 1, "", "get_wu_pre_var_location"], [7, 2, 1, "", "get_wu_var_location"], [7, 3, 1, "", "kernel_size"], [7, 3, 1, "", "matrix_type"], [7, 3, 1, "", "max_connections"], [7, 3, 1, "", "max_dendritic_delay_timesteps"], [7, 3, 1, "", "max_source_connections"], [7, 3, 1, "", "name"], [7, 3, 1, "", "narrow_sparse_ind_enabled"], [7, 3, 1, "", "num_threads_per_spike"], [7, 3, 1, "", "output_location"], [7, 3, 1, "", "parallelism_hint"], [7, 3, 1, "", "post_target_var"], [7, 3, 1, "", "pre_target_var"], [7, 3, 1, "", "ps_initialiser"], [7, 2, 1, "", "set_ps_param_dynamic"], [7, 2, 1, "", "set_ps_var_location"], [7, 2, 1, "", "set_wu_param_dynamic"], [7, 2, 1, "", "set_wu_post_var_location"], [7, 2, 1, "", "set_wu_pre_var_location"], [7, 2, 1, "", "set_wu_var_location"], [7, 3, 1, "", "sparse_connectivity_initialiser"], [7, 3, 1, "", "sparse_connectivity_location"], [7, 3, 1, "", "toeplitz_connectivity_initialiser"], [7, 3, 1, "", "wu_initialiser"]], "pygenn.SynapseMatrixConnectivity": [[7, 4, 1, "", "BITMASK"], [7, 4, 1, "", "DENSE"], [7, 4, 1, "", "PROCEDURAL"], [7, 4, 1, "", "SPARSE"], [7, 4, 1, "", "TOEPLITZ"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.SynapseMatrixType": [[7, 4, 1, "", "BITMASK"], [7, 4, 1, "", "DENSE"], [7, 4, 1, "", "DENSE_PROCEDURALG"], [7, 4, 1, "", "PROCEDURAL"], [7, 4, 1, "", "PROCEDURAL_KERNELG"], [7, 4, 1, "", "SPARSE"], [7, 4, 1, "", "TOEPLITZ"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.SynapseMatrixWeight": [[7, 4, 1, "", "INDIVIDUAL"], [7, 4, 1, "", "KERNEL"], [7, 4, 1, "", "PROCEDURAL"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.VarAccess": [[7, 4, 1, "", "READ_ONLY"], [7, 4, 1, "", "READ_ONLY_DUPLICATE"], [7, 4, 1, "", "READ_ONLY_SHARED_NEURON"], [7, 4, 1, "", "READ_WRITE"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.VarAccessDim": [[7, 4, 1, "", "BATCH"], [7, 4, 1, "", "ELEMENT"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.VarAccessMode": [[7, 4, 1, "", "READ_ONLY"], [7, 4, 1, "", "READ_WRITE"], [7, 4, 1, "", "REDUCE_MAX"], [7, 4, 1, "", "REDUCE_SUM"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.VarAccessModeAttribute": [[7, 4, 1, "", "MAX"], [7, 4, 1, "", "READ_ONLY"], [7, 4, 1, "", "READ_WRITE"], [7, 4, 1, "", "REDUCE"], [7, 4, 1, "", "SUM"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.VarLocation": [[7, 4, 1, "", "DEVICE"], [7, 4, 1, "", "HOST_DEVICE"], [7, 4, 1, "", "HOST_DEVICE_ZERO_COPY"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.VarLocationAttribute": [[7, 4, 1, "", "DEVICE"], [7, 4, 1, "", "HOST"], [7, 4, 1, "", "ZERO_COPY"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.cuda_backend": [[7, 1, 1, "", "BlockSizeSelect"], [7, 1, 1, "", "DeviceSelect"], [7, 1, 1, "", "Preferences"]], "pygenn.cuda_backend.BlockSizeSelect": [[7, 4, 1, "", "MANUAL"], [7, 4, 1, "", "OCCUPANCY"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.cuda_backend.DeviceSelect": [[7, 4, 1, "", "MANUAL"], [7, 4, 1, "", "MOST_MEMORY"], [7, 4, 1, "", "OPTIMAL"], [7, 3, 1, "", "name"], [7, 3, 1, "", "value"]], "pygenn.cuda_backend.Preferences": [[7, 3, 1, "", "block_size_select_method"], [7, 3, 1, "", "constant_cache_overhead"], [7, 3, 1, "", "device_select_method"], [7, 3, 1, "", "enable_nccl_reductions"], [7, 3, 1, "", "generate_line_info"], [7, 3, 1, "", "manual_block_sizes"], [7, 3, 1, "", "manual_device_id"], [7, 3, 1, "", "show_ptx_info"]], "pygenn.current_source_models": [[7, 5, 1, "", "DC"], [7, 5, 1, "", "GaussianNoise"], [7, 5, 1, "", "PoissonExp"]], "pygenn.custom_update_models": [[7, 5, 1, "", "Transpose"]], "pygenn.genn_groups": [[7, 1, 1, "", "CurrentSourceMixin"], [7, 1, 1, "", "CustomConnectivityUpdateMixin"], [7, 1, 1, "", "CustomUpdateMixin"], [7, 1, 1, "", "CustomUpdateWUMixin"], [7, 1, 1, "", "GroupMixin"], [7, 1, 1, "", "NeuronGroupMixin"], [7, 1, 1, "", "SynapseGroupMixin"]], "pygenn.genn_groups.CustomConnectivityUpdateMixin": [[7, 2, 1, "", "get_var_values"]], "pygenn.genn_groups.CustomUpdateWUMixin": [[7, 2, 1, "", "get_var_values"]], "pygenn.genn_groups.GroupMixin": [[7, 2, 1, "", "pull_extra_global_param_from_device"], [7, 2, 1, "", "pull_var_from_device"], [7, 2, 1, "", "push_extra_global_param_to_device"], [7, 2, 1, "", "push_var_to_device"], [7, 2, 1, "", "set_dynamic_param_value"]], "pygenn.genn_groups.NeuronGroupMixin": [[7, 3, 1, "", "spike_recording_data"]], "pygenn.genn_groups.SynapseGroupMixin": [[7, 2, 1, "", "get_sparse_post_inds"], [7, 2, 1, "", "get_sparse_pre_inds"], [7, 2, 1, "", "get_var_values"], [7, 3, 1, "", "post_spike_event_recording_data"], [7, 3, 1, "", "pre_spike_event_recording_data"], [7, 2, 1, "", "pull_connectivity_from_device"], [7, 2, 1, "", "pull_in_syn_from_device"], [7, 2, 1, "", "pull_psm_extra_global_param_from_device"], [7, 2, 1, "", "push_connectivity_to_device"], [7, 2, 1, "", "push_in_syn_to_device"], [7, 2, 1, "", "push_psm_extra_global_param_to_device"], [7, 2, 1, "", "set_sparse_connections"], [7, 3, 1, "", "synapse_group"], [7, 3, 1, "", "weight_update_var_size"]], "pygenn.init_sparse_connectivity_snippets": [[7, 5, 1, "", "Conv2D"], [7, 5, 1, "", "FixedNumberPostWithReplacement"], [7, 5, 1, "", "FixedNumberPreWithReplacement"], [7, 5, 1, "", "FixedNumberTotalWithReplacement"], [7, 5, 1, "", "FixedProbability"], [7, 5, 1, "", "FixedProbabilityNoAutapse"], [7, 5, 1, "", "OneToOne"], [7, 5, 1, "", "Uninitialised"]], "pygenn.init_toeplitz_connectivity_snippets": [[7, 5, 1, "", "AvgPoolConv2D"], [7, 5, 1, "", "Conv2D"], [7, 5, 1, "", "Uninitialised"]], "pygenn.init_var_snippets": [[7, 5, 1, "", "Binomial"], [7, 5, 1, "", "Constant"], [7, 5, 1, "", "Exponential"], [7, 5, 1, "", "Gamma"], [7, 5, 1, "", "Kernel"], [7, 5, 1, "", "Normal"], [7, 5, 1, "", "NormalClipped"], [7, 5, 1, "", "NormalClippedDelay"], [7, 5, 1, "", "Uniform"], [7, 5, 1, "", "Uninitialised"]], "pygenn.model_preprocessor": [[7, 1, 1, "", "Array"], [7, 1, 1, "", "ArrayBase"], [7, 1, 1, "", "ExtraGlobalParameter"], [7, 1, 1, "", "SynapseVariable"], [7, 1, 1, "", "Variable"], [7, 1, 1, "", "VariableBase"]], "pygenn.model_preprocessor.Array": [[7, 3, 1, "", "view"]], "pygenn.model_preprocessor.ArrayBase": [[7, 2, 1, "", "pull_from_device"], [7, 2, 1, "", "push_to_device"], [7, 2, 1, "", "set_array"]], "pygenn.model_preprocessor.ExtraGlobalParameter": [[7, 2, 1, "", "set_init_values"], [7, 2, 1, "", "set_values"], [7, 3, 1, "", "values"], [7, 3, 1, "", "view"]], "pygenn.model_preprocessor.SynapseVariable": [[7, 3, 1, "", "current_values"], [7, 3, 1, "", "current_view"], [7, 3, 1, "", "values"], [7, 3, 1, "", "view"]], "pygenn.model_preprocessor.Variable": [[7, 3, 1, "", "current_values"], [7, 3, 1, "", "current_view"], [7, 3, 1, "", "values"], [7, 3, 1, "", "view"]], "pygenn.model_preprocessor.VariableBase": [[7, 2, 1, "", "set_array"], [7, 2, 1, "", "set_init_values"], [7, 2, 1, "", "set_values"]], "pygenn.neuron_models": [[7, 5, 1, "", "Izhikevich"], [7, 5, 1, "", "IzhikevichVariable"], [7, 5, 1, "", "LIF"], [7, 5, 1, "", "Poisson"], [7, 5, 1, "", "PoissonNew"], [7, 5, 1, "", "RulkovMap"], [7, 5, 1, "", "SpikeSource"], [7, 5, 1, "", "SpikeSourceArray"], [7, 5, 1, "", "TraubMiles"], [7, 5, 1, "", "TraubMilesAlt"], [7, 5, 1, "", "TraubMilesFast"], [7, 5, 1, "", "TraubMilesNStep"]], "pygenn.postsynaptic_models": [[7, 5, 1, "", "DeltaCurr"], [7, 5, 1, "", "ExpCond"], [7, 5, 1, "", "ExpCurr"]], "pygenn.single_threaded_cpu_backend": [[7, 1, 1, "", "Preferences"]], "pygenn.weight_update_models": [[7, 5, 1, "", "PiecewiseSTDP"], [7, 5, 1, "", "StaticGraded"], [7, 5, 1, "", "StaticPulse"], [7, 5, 1, "", "StaticPulseConstantWeight"], [7, 5, 1, "", "StaticPulseDendriticDelay"]]}, "objtypes": {"0": "py:module", "1": "py:class", "2": "py:method", "3": "py:property", "4": "py:attribute", "5": "py:function"}, "objnames": {"0": ["py", "module", "Python module"], "1": ["py", "class", "Python class"], "2": ["py", "method", "Python method"], "3": ["py", "property", "Python property"], "4": ["py", "attribute", "Python attribute"], "5": ["py", "function", "Python function"]}, "titleterms": {"bibliographi": 0, "build": 1, "network": [1, 5], "The": 1, "model": [1, 2, 10, 11], "popul": 1, "paramet": [1, 5], "extra": [1, 5], "global": [1, 5], "refer": 1, "variabl": [1, 2, 5], "locat": 1, "neuron": [1, 2], "synaps": 1, "current": [1, 2], "sourc": [1, 2], "custom": [1, 2], "updat": [1, 2], "connect": [1, 2], "genncod": [2, 8], "random": 2, "number": 2, "gener": 2, "initialis": 2, "snippet": 2, "spars": 2, "toeplitz": 2, "weight": 2, "postsynapt": 2, "pygenn": [3, 6, 7, 11, 13], "document": 3, "indic": 3, "tabl": 3, "comput": [4, 12], "time": [4, 12], "simul": 5, "spike": 5, "record": 5, "push": 5, "pull": 5, "valu": 5, "view": 5, "dynam": 5, "packag": 7, "submodul": 7, "cuda_backend": 7, "modul": 7, "current_source_model": 7, "custom_connectivity_update_model": 7, "custom_update_model": 7, "genn_group": 7, "init_sparse_connectivity_snippet": 7, "init_toeplitz_connectivity_snippet": 7, "init_var_snippet": 7, "model_preprocessor": 7, "neuron_model": 7, "postsynaptic_model": 7, "runtim": 7, "single_threaded_cpu_backend": 7, "type": 7, "weight_update_model": 7, "upgrad": 8, "from": 8, "genn": 8, "4": 8, "syntax": 8, "chang": 8, "user": 9, "project": 9, "mnist": 10, "classif": 10, "us": 10, "an": 10, "insect": 10, "inspir": 10, "mushroom": 10, "bodi": 10, "name": [10, 11, 13], "argument": [10, 11, 13], "implement": [11, 13], "local": 11, "cortic": 11, "microcircuit": 11, "superspik": 13}, "envversion": {"sphinx.domains.c": 2, "sphinx.domains.changeset": 1, "sphinx.domains.citation": 1, "sphinx.domains.cpp": 6, "sphinx.domains.index": 1, "sphinx.domains.javascript": 2, "sphinx.domains.math": 2, "sphinx.domains.python": 3, "sphinx.domains.rst": 2, "sphinx.domains.std": 2, "sphinx": 56}})
\ No newline at end of file
+Search.setIndex({"docnames": ["bibliography", "building_networks", "custom_models", "index", "installation", "sg_execution_times", "simulating_networks", "source/modules", "source/pygenn", "upgrading", "userproject/index", "userproject/mnist_mb_classifier", "userproject/potjans_microcircuit", "userproject/sg_execution_times", "userproject/superspike_demo"], "filenames": ["bibliography.rst", "building_networks.rst", "custom_models.rst", "index.rst", "installation.rst", "sg_execution_times.rst", "simulating_networks.rst", "source\\modules.rst", "source\\pygenn.rst", "upgrading.rst", "userproject\\index.rst", "userproject\\mnist_mb_classifier.rst", "userproject\\potjans_microcircuit.rst", "userproject\\sg_execution_times.rst", "userproject\\superspike_demo.rst"], "titles": ["Bibliography", "Building networks", "Custom models", "PyGeNN documentation", "Installation", "Computation times", "Simulating networks", "pygenn", "pygenn package", "Upgrading from GeNN 4", "User projects", "MNIST classification using an insect-inspired mushroom body model", "PyGeNN implementation of local cortical microcircuit model", "Computation times", "PyGeNN implementation of SuperSpike"], "terms": {"morrison2008": [0, 2, 8], "morrison": 0, "A": [0, 1, 2, 8], "diesmann": [0, 12], "m": [0, 1, 8, 14], "gerstner": 0, "w": [0, 2, 14], "2008": 0, "phenomenolog": 0, "model": [0, 3, 5, 6, 8, 9, 10, 13, 14], "synapt": [0, 1, 2, 8, 11, 12, 14], "plastic": 0, "base": [0, 1, 2, 6, 8, 14], "spike": [0, 1, 2, 3, 8, 11, 12, 14], "time": [0, 1, 2, 6, 8, 9, 11, 12, 14], "biolog": 0, "cybernet": 0, "98": [0, 8], "459": 0, "478": 0, "http": [0, 2, 4], "doi": 0, "org": [0, 4], "10": [0, 1, 6, 8, 11, 12, 14], "1007": 0, "s00422": 0, "008": [0, 12], "0233": [0, 11], "1": [0, 1, 2, 3, 6, 8, 11, 12, 14], "potjans2014": [0, 12], "potjan": [0, 12], "t": [0, 2, 4, 6, 8, 9, 11, 12, 14], "c": [0, 1, 2, 3, 4, 8, 11, 12, 14], "2014": 0, "The": [0, 2, 3, 4, 6, 8, 9, 12], "cell": [0, 11, 12], "type": [0, 1, 2, 7, 9, 12, 14], "specif": [0, 12], "cortic": [0, 5, 10, 13], "microcircuit": [0, 5, 10, 13], "relat": [0, 1, 8], "structur": [0, 1, 2, 8, 9, 14], "activ": [0, 2, 8, 12], "full": [0, 2, 8, 11, 12, 14], "scale": [0, 8, 11, 12], "network": [0, 2, 3, 8, 11, 12, 14], "cerebr": 0, "cortex": 0, "24": 0, "3": [0, 1, 5, 8, 11, 12, 13, 14], "785": 0, "806": 0, "1093": 0, "cercor": 0, "bhs358": 0, "zenke2018": [0, 14], "zenk": [0, 14], "f": [0, 11, 12, 14], "ganguli": [0, 14], "s": [0, 1, 2, 6, 8, 11, 12, 14], "2018": 0, "superspik": [0, 5, 10, 13], "supervis": [0, 11], "learn": [0, 1, 2, 3, 8, 11, 14], "multilay": 0, "neural": [0, 3], "comput": [0, 2, 8, 14], "30": [0, 2, 8], "6": [0, 1, 8, 12], "1514": 0, "1541": 0, "1162": 0, "neco_a_01086": 0, "knight2018": [0, 2, 8], "knight": [0, 3], "j": 0, "nowotni": [0, 3], "gpu": [0, 1, 2, 3, 4, 6, 8, 9, 11], "outperform": 0, "current": [0, 3, 6, 8, 9, 11, 12], "hpc": 0, "neuromorph": 0, "solut": 0, "term": [0, 1, 2, 8], "speed": [0, 1, 8], "energi": 0, "when": [0, 1, 2, 4, 6, 8, 9, 11], "simul": [0, 1, 2, 3, 8, 11, 12], "highli": 0, "connect": [0, 8, 9, 11, 12], "frontier": 0, "neurosci": [0, 1], "12": [0, 8], "decemb": 0, "19": 0, "3389": 0, "fnin": 0, "00941": 0, "turner2022": [0, 2, 8], "turner": 0, "p": [0, 2, 8, 12], "subramanian": 0, "2022": 0, "mlgenn": 0, "acceler": [0, 1, 3], "snn": [0, 2], "infer": 0, "us": [0, 1, 2, 3, 4, 5, 6, 8, 9, 10, 12, 13, 14], "enabl": [0, 1, 6, 12], "engin": 0, "2": [0, 1, 2, 8, 11, 12, 14], "024002": 0, "1088": 0, "2634": 0, "4386": 0, "ac5ac5": 0, "defin": [1, 2, 8, 11], "follow": [1, 2, 6, 8, 11, 12, 14], "gennmodel": [1, 2, 6, 8, 9, 11, 12, 14], "must": [1, 2, 8], "creat": [1, 2, 4, 8, 9, 11, 12, 14], "default": [1, 2, 6, 8, 11, 12, 14], "precis": [1, 2, 8, 12], "see": [1, 2, 8, 9, 12], "ref": 1, "floatprecis": 1, "name": [1, 2, 6, 8], "float": [1, 2, 8, 11, 12, 14], "yourmodelnam": 1, "By": [1, 2, 6, 8], "hardwar": [1, 2, 3, 8, 9], "code": [1, 2, 3, 6, 8, 9, 10, 11, 12, 14], "gener": [1, 3, 6, 8, 10, 11, 12, 14], "backend": [1, 6, 8], "avail": [1, 8], "howev": [1, 2, 4, 6, 8], "thi": [1, 2, 3, 4, 6, 8, 9, 11, 12, 14], "can": [1, 2, 4, 6, 8, 9, 11, 12, 14], "overriden": 1, "keyword": [1, 2, 6, 8], "argument": [1, 2, 6, 8], "For": [1, 2, 4, 6, 8, 9], "exampl": [1, 2, 4, 5, 6, 8, 9, 10, 11, 12, 13, 14], "singl": [1, 2, 6, 8, 9], "thread": [1, 2, 6, 8, 12], "cpu": [1, 2, 6, 8], "could": [1, 2, 6, 8, 9], "manual": [1, 4, 8], "select": [1, 4, 8], "single_threaded_cpu": 1, "run": [1, 2, 4, 6, 8], "smaller": 1, "mai": [1, 2, 8], "fulli": [1, 2, 8], "occupi": 1, "devic": [1, 2, 6, 8, 11, 12], "In": [1, 2, 4, 6, 8, 9, 12], "some": [1, 2, 8], "scenario": [1, 2, 8], "gradient": [1, 2, 8, 14], "train": [1, 11, 14], "sweep": 1, "overcom": 1, "rune": 1, "multipl": [1, 2, 8], "copi": [1, 2, 6, 8, 11], "same": [1, 2, 6, 8, 9], "batch": [1, 2, 6, 8], "machin": [1, 3, 4], "speak": [1, 8], "genn": [1, 2, 3, 4, 6, 7, 10, 14], "batch_siz": 1, "512": 1, "spars": [1, 6, 8, 9, 11, 12, 14], "ar": [1, 2, 3, 4, 6, 8, 9, 12, 14], "share": [1, 2, 6, 8], "across": [1, 2, 6, 8], "all": [1, 2, 5, 6, 8, 9, 10, 12, 14], "whether": [1, 2, 8], "state": [1, 2, 6, 8, 9, 11], "duplic": [1, 2, 8], "control": 1, "varaccess": [1, 2, 8], "customupdatevaraccess": [1, 2, 8], "associ": [1, 2, 6, 8], "each": [1, 2, 4, 6, 8, 11, 14], "pleas": [1, 2, 3, 8], "todo": [1, 14], "more": [1, 2, 9, 12], "detail": [1, 2, 8], "addition": [1, 2, 6, 8], "ani": [1, 2, 6, 8], "prefer": [1, 8], "expos": [1, 2, 8], "configur": [1, 2, 4, 8], "here": [1, 2, 8], "cuda": [1, 4, 6, 8], "allow": [1, 2, 8, 9, 11, 12], "you": [1, 2, 3, 4, 6, 8, 9], "which": [1, 2, 3, 4, 6, 8, 9], "via": [1, 2, 4, 6, 8, 9], "manual_device_id": [1, 8], "0": [1, 2, 5, 6, 8, 11, 12, 13, 14], "formalis": 1, "concept": 1, "group": [1, 6, 8, 9], "function": [1, 2, 6, 8, 9, 12, 14], "practic": 1, "e": [1, 2, 4, 6, 8, 12, 14], "g": [1, 2, 4, 6, 8, 11, 12], "brain": [1, 3], "region": 1, "layer": [1, 12], "context": [1, 2, 8], "initialis": [1, 8, 9, 14], "constant": [1, 2, 8, 11], "numer": [1, 2, 8], "valu": [1, 2, 8, 11, 12], "homogen": [1, 2], "an": [1, 2, 4, 5, 6, 8, 10, 13], "entir": 1, "ini": 1, "0529324": 1, "thei": [1, 2, 6, 8, 9], "veri": [1, 2, 6, 8], "effici": [1, 2, 8, 9], "access": [1, 2, 6, 8, 9], "from": [1, 2, 3, 4, 5, 6, 8, 11, 12, 13, 14], "either": [1, 4, 8], "hard": 1, "deliv": [1, 2, 8, 9, 11], "high": [1, 8], "perform": [1, 2, 8], "cach": [1, 8], "onli": [1, 2, 8, 9, 11], "liter": [1, 2], "chang": [1, 4, 6, 8], "need": [1, 2, 4, 6, 8, 9], "member": [1, 8], "have": [1, 2, 4, 6, 8, 9, 14], "exact": 1, "often": [1, 6], "b": [1, 6, 8], "contain": [1, 2, 3, 8], "individu": [1, 8], "over": [1, 2, 6], "mani": [1, 2, 8, 11, 12], "wai": [1, 2, 8, 9], "through": [1, 2, 8, 12, 14], "python": [1, 3, 4, 6, 10, 11, 12, 14], "dictionari": [1, 6, 12], "pass": [1, 2, 8, 9], "add_neuron_popul": [1, 8, 11, 12, 14], "add_synapse_popul": [1, 8, 9, 11, 12, 14], "To": [1, 6, 9], "one": [1, 2, 4, 8], "fill": 1, "them": [1, 2, 6, 8], "sequenc": [1, 2, 8], "np": [1, 2, 6, 11, 12, 14], "arang": [1, 6, 12], "400": [1, 14], "snippet": [1, 3, 8, 9], "return": [1, 2, 8, 11, 12, 14], "pygenn": [1, 2, 5, 6, 9, 10, 11, 13], "init_var": [1, 8, 12, 14], "param": [1, 2, 8, 11, 14], "union": [1, 2, 8], "initvarsnippetbas": [1, 8], "str": [1, 2, 8, 14], "init": [1, 8, 9, 12, 14], "string": [1, 2, 3, 8, 9], "referenc": [1, 2, 8, 9], "built": [1, 2, 4, 6, 8], "init_var_snippet": [1, 7], "instanc": [1, 8], "create_var_init_snippet": [1, 2, 8], "dict": [1, 8], "int": [1, 2, 8, 9, 11, 12, 14], "normal": [1, 2, 8, 12], "sampl": [1, 2, 8], "distribut": [1, 2, 4, 8, 11, 12], "mean": [1, 2, 8, 12, 14], "standard": [1, 2, 4, 8, 11, 12], "deviat": [1, 2, 8, 12], "sd": [1, 2, 8, 12, 14], "result": [1, 8], "usual": [1, 8], "As": [1, 2, 6, 8], "well": [1, 2, 6, 8], "variou": [1, 2], "belong": 1, "other": [1, 2, 4, 8], "postsynapt": [1, 6, 8, 9, 12], "attach": [1, 2, 8], "call": [1, 2, 3, 6, 8, 9], "r": [1, 2, 8, 14], "assign": [1, 2, 8], "syntax": 1, "neuron_var_ref": [1, 2, 8], "create_var_ref": [1, 8, 14], "ng": 1, "v": [1, 2, 6, 8, 9, 11, 12, 14], "current_source_var_ref": 1, "cs": [1, 8], "custom_update_var_ref": 1, "cu": 1, "postsynaptic_model_var_ref": 1, "create_psm_var_ref": [1, 8], "sg": [1, 8], "wu_pre_var_ref": 1, "create_wu_pre_var_ref": [1, 8], "pre": [1, 2, 8], "wu_post_var_ref": 1, "create_wu_post_var_ref": [1, 8], "post": [1, 2, 8], "where": [1, 2, 6, 8, 9], "neurongroup": [1, 6, 8], "currentsourc": [1, 2, 8], "add_current_sourc": [1, 8, 11, 12], "customupd": [1, 8], "add_custom_upd": [1, 8, 14], "synapsegroup": [1, 2, 6, 8, 9], "while": [1, 2, 6, 8, 12, 14], "interchang": 1, "long": [1, 2, 11], "delai": [1, 2, 6, 8, 9, 12], "size": [1, 2, 8, 9, 11, 14], "per": [1, 2, 6, 8, 12, 14], "weight": [1, 6, 8, 9, 11, 12, 14], "slightli": 1, "differ": [1, 2, 6, 8, 12], "wu_var_ref": 1, "create_wu_var_ref": [1, 8, 14], "cu_wu_var_ref": 1, "customupdatewu": [1, 8], "oper": [1, 2, 6, 8, 9], "anoth": [1, 8], "These": [1, 2, 8], "also": [1, 2, 4, 6, 8], "addit": [1, 2, 8], "featur": [1, 2, 9], "link": 1, "transpos": [1, 8, 14], "wu_transpose_var_ref": 1, "back_sg": 1, "tranpos": [1, 8, 14], "dimens": [1, 8], "i": [1, 2, 4, 8, 9, 11, 12, 14], "its": [1, 8], "_postsynaptic_": 1, "ha": [1, 2, 4, 6, 8, 9], "number": [1, 3, 6, 8, 9, 11, 12, 14], "_presynaptic_": [1, 2, 8], "vice": 1, "versa": 1, "after": [1, 2, 6, 8, 11], "made": [1, 6, 8, 9], "forward": [1, 2, 8], "appli": [1, 2, 6, 8, 12], "_": [1, 2, 6, 8, 9, 12], "possibl": [1, 2, 8, 9], "synapsematrixtyp": [1, 2, 8, 9], "dens": [1, 8, 9, 11, 14], "onc": [1, 2, 6, 8, 11], "how": [1, 2, 8, 11, 12], "your": [1, 2, 4, 6, 8], "go": [1, 11, 12, 14], "alloc": [1, 6], "memori": [1, 2, 6, 8, 11], "both": [1, 6, 8], "host": [1, 2, 6, 8], "altern": [1, 4], "class": [1, 2, 6, 8], "varloc": [1, 6, 8, 12], "support": [1, 2, 3, 8, 9], "combin": [1, 2, 8], "varlocationattribut": [1, 8], "save": [1, 6, 8, 11, 12, 14], "host_devic": [1, 8], "host_device_zero_copi": [1, 8], "between": [1, 2, 6, 8, 12, 14], "zero": [1, 2, 6, 8, 14], "improv": [1, 8, 9], "data": [1, 2, 6, 8, 11, 12, 14], "frequent": [1, 8], "non": [1, 2, 8, 9], "coher": [1, 8], "architectur": [1, 8], "jetson": [1, 8], "reduc": [1, 2, 8, 9, 14], "note": [1, 3, 8, 12, 14], "newer": 1, "embed": 1, "system": [1, 6, 8, 9], "tx1": 1, "physic": 1, "seper": [1, 8, 9], "thu": [1, 6], "ad": [1, 2, 4, 8, 9], "pop_nam": [1, 8, 12], "num_neuron": [1, 2, 8, 12], "var": [1, 2, 6, 8, 11, 14], "add": [1, 2, 8, 9, 12, 14], "uniqu": [1, 8, 12], "neuronmodelbas": [1, 8], "neuron_model": [1, 7], "create_neuron_model": [1, 2, 8, 11, 14], "varinit": [1, 8], "ndarrai": [1, 8], "initi": [1, 2, 6, 8, 11, 14], "izhikevich": [1, 8], "set": [1, 2, 4, 6, 8, 9, 11, 12], "tonic": [1, 8], "pop": [1, 6, 8, 11, 12], "02": [1, 2, 8], "65": [1, 8, 12], "d": [1, 8, 12, 14], "u": [1, 8, 12, 14], "20": [1, 8, 11], "two": [1, 2, 8, 9], "Their": 1, "behaviour": [1, 2, 3, 8], "describ": [1, 2, 3, 6, 8, 9], "what": [1, 2, 3, 8], "kind": 1, "dynam": [1, 2, 8, 9, 14], "occur": [1, 2, 8], "output": [1, 2, 8, 11, 12, 14], "presynapt": [1, 2, 8, 9, 14], "typic": [1, 2, 8], "init_weight_upd": [1, 8, 9, 11, 12, 14], "pre_var": [1, 2, 6, 8, 14], "post_var": [1, 2, 6, 8, 14], "pre_var_ref": [1, 2, 8, 14], "post_var_ref": [1, 2, 8, 14], "weight_update_model": [1, 7], "weightupdatemodelbas": [1, 8], "create_weight_update_model": [1, 2, 8, 11, 14], "varrefer": [1, 8], "static": [1, 8], "puls": [1, 8], "weight_init": [1, 8], "staticpulseconstantweight": [1, 8, 9, 11], "input": [1, 2, 8, 9, 11, 12, 14], "translat": [1, 2], "init_postsynapt": [1, 8, 9, 11, 12, 14], "var_ref": [1, 2, 8, 14], "postsynapticmodelbas": [1, 8], "postsynaptic_models_model": [1, 8], "create_postsynaptic_model": [1, 2, 8, 14], "conduct": [1, 8], "exponenti": [1, 2, 8, 14], "shape": [1, 2, 8, 11], "postsynaptic_init": [1, 8], "expcond": [1, 8], "tau": [1, 2, 6, 8, 11, 12, 14], "80": [1, 8, 11, 14], "pop1": [1, 8], "provid": [1, 2, 4, 6, 8], "implement": [1, 2, 5, 8, 9, 10, 13], "dense_proceduralg": [1, 8, 9], "bitmask": [1, 2, 8], "procedur": [1, 2, 8, 9, 12], "procedural_kernelg": [1, 8, 9], "toeplitz": [1, 8, 9], "dense_procedur": 1, "simpli": [1, 8], "init_sparse_connect": [1, 8, 11, 12], "initsparseconnectivitysnippetbas": [1, 8], "init_sparse_connectivity_snippet": [1, 7], "create_sparse_connect_init_snippet": [1, 2, 8], "pair": [1, 2, 8], "probabl": [1, 8, 12], "fixedprob": [1, 8], "prob": [1, 8], "init_toeplitz_connect": [1, 8], "init_toeplitz_connect_snippet": [1, 8], "init_toeplitz_connectivity_snippet": [1, 7], "inittoeplitzconnectivitysnippetbas": [1, 8], "create_toeplitz_connect_init_snippet": [1, 2, 8], "2d": [1, 8], "convolut": [1, 2, 8], "kernel": [1, 2, 6, 8, 12, 14], "64": [1, 2, 8], "62": [1, 8], "conv_kh": [1, 8], "conv_kw": [1, 8], "conv_ih": [1, 8], "conv_iw": [1, 8], "conv_ic": [1, 8], "conv_oh": [1, 8], "conv_ow": [1, 8], "conv_oc": [1, 8], "conv2d": [1, 8], "should": [1, 2, 4, 8, 11], "4096": [1, 8], "3844": [1, 8], "final": [1, 2, 8], "compon": 1, "place": [1, 6, 8], "matrix_typ": [1, 8, 12], "target": [1, 2, 8, 9, 12, 14], "weight_update_init": [1, 8], "connectivity_init": [1, 8], "none": [1, 2, 8, 11, 12, 14], "sparseconnectivityinit": [1, 8], "toeplitzconnectivityinit": [1, 8], "init_toeplitz_connectivity_connect": [1, 8], "src_pop": [1, 8, 12], "target_pop": [1, 8], "syn": [1, 2, 8], "expcurr": [1, 8, 11, 12, 14], "5": [1, 4, 8, 9, 11, 12, 14], "cs_name": [1, 8], "current_source_model": [1, 7], "currentsourcemodelbas": [1, 8], "create_current_source_model": [1, 2, 8, 11], "inject": [1, 2, 8, 11], "gaussian": [1, 8], "nois": [1, 2, 8, 14], "gaussiannois": [1, 8], "previou": [1, 2, 6, 8, 9, 14], "section": 1, "automat": [1, 2, 4, 8, 12], "everi": [1, 2, 8, 12, 14], "timestep": [1, 2, 6, 8, 11, 12, 14], "process": [1, 6, 9], "would": [1, 2, 6, 8, 9], "benefit": 1, "trigger": [1, 2, 8], "occasion": 1, "classifi": 1, "reset": [1, 2, 8, 11, 14], "stimuli": 1, "been": [1, 8, 9], "present": [1, 4, 11], "optim": [1, 8], "accumul": [1, 2, 8], "sever": [1, 2, 8, 9], "similar": [1, 9], "preced": [1, 8], "calcul": [1, 2, 8, 12, 14], "cu_nam": [1, 8], "group_nam": [1, 8], "custom_update_model": [1, 7], "egp_ref": [1, 8], "includ": [1, 2, 3, 8, 12], "execut": [1, 2, 5, 8, 13], "simultan": [1, 8], "customupdatemodelbas": [1, 8], "create_custom_update_model": [1, 2, 8, 14], "wuvarrefer": [1, 8], "egprefer": [1, 8], "create_egp_ref": [1, 8], "like": [1, 2, 3, 6, 8, 9], "user": [1, 2, 3, 6, 8, 9], "rather": [1, 2, 8, 9, 11, 12, 14], "than": [1, 2, 8, 11, 12, 14], "add_custom_connectivity_upd": [1, 8], "syn_group": [1, 8], "custom_conn_update_model": [1, 8], "customconnectivityupdatemodelbas": [1, 8], "custom_connectivity_update_model": [1, 7], "customconnectivityupdatemodelbaseupdatemodelbas": [1, 8], "create_custom_connectivity_update_model": [1, 2, 8], "One": [2, 6, 8], "main": [2, 8], "thing": [2, 9], "make": [2, 3, 4, 6, 8, 14], "build": [2, 3, 6, 8, 9, 11, 12, 14], "easili": [2, 3, 8], "customis": [2, 3], "languag": [2, 3, 9], "we": [2, 4, 6, 8, 9, 11], "essenti": [2, 8, 14], "c99": [2, 9], "en": [2, 4], "cpprefer": 2, "com": [2, 4], "No": [2, 8], "preprocessor": 2, "enough": 2, "printf": 2, "debug": [2, 4, 8], "messag": 2, "much": [2, 6, 8, 9], "strstr": 2, "etc": [2, 14], "typedefin": 2, "cannot": [2, 3, 6, 8], "esoter": 2, "octal": 2, "integ": [2, 6, 8], "hexadecim": 2, "point": [2, 6, 8, 12, 14], "aren": 2, "address": [2, 9], "isn": [2, 9], "On": [2, 4, 6], "local": [2, 4, 5, 8, 9, 10, 13], "assum": [2, 4, 9], "store": [2, 8, 9], "regist": [2, 9], "limit": [2, 9], "deal": [2, 9], "extra": [2, 8, 9], "global": [2, 8, 9], "paramet": [2, 8, 9, 11, 12, 14], "arrai": [2, 8, 9, 14], "longer": [2, 8, 9], "do": [2, 4, 6, 8, 9], "stuff": [2, 9], "const": [2, 8, 9, 11, 14], "egpsubset": [2, 9], "egp": [2, 9], "offset": [2, 8, 9, 12], "instead": [2, 8, 9], "overload": [2, 8], "so": [2, 6, 8, 9, 14], "sin": 2, "0f": 2, "resolv": 2, "doubl": 2, "version": [2, 4, 8, 9], "without": 2, "suffix": [2, 8], "treat": [2, 8], "scalar": [2, 8, 11, 14], "alwai": 2, "0d": 2, "lp64": 2, "platform": [2, 6], "32": [2, 8, 14], "bit": 2, "librari": [2, 4, 8], "co": 2, "tan": 2, "aco": 2, "asin": 2, "atan": 2, "atan2": 2, "cosh": 2, "sinh": 2, "tanh": [2, 8], "acosh": 2, "asinh": 2, "atanh": 2, "exp": [2, 8, 11, 14], "expm1": 2, "exp2": 2, "pow": [2, 14], "scalbn": 2, "log": [2, 8, 12, 14], "log1p": 2, "log2": 2, "log10": 2, "ldexp": 2, "ilogb": 2, "sqrt": [2, 12, 14], "cbrt": 2, "hypot": 2, "ceil": [2, 8], "floor": 2, "fmod": 2, "round": [2, 11, 12], "rint": 2, "trunc": 2, "nearbyint": 2, "nextaft": 2, "remaind": 2, "fab": [2, 14], "fdim": 2, "fmax": [2, 8, 11, 14], "fmin": [2, 8, 11, 14], "erf": 2, "erfc": 2, "tgamma": 2, "lgamma": 2, "copysign": 2, "fma": 2, "min": [2, 8, 12, 14], "max": [2, 8, 11, 12, 14], "ab": [2, 12], "form": [2, 8], "probabilist": 2, "mechan": [2, 8, 9], "within": [2, 8, 14], "gennrand_uniform": [2, 8], "drawn": [2, 8], "uniformli": [2, 8], "interv": 2, "gennrand_norm": [2, 8], "gennrand_exponenti": 2, "lambda": [2, 8, 14], "gennrand_log_norm": 2, "std": 2, "specifi": [2, 8], "gennrand_gamma": 2, "alpha": [2, 8, 14], "gamma": [2, 8], "gennrand_binomi": 2, "n": [2, 8, 11, 12], "binomi": [2, 8], "part": [2, 8], "deriv": [2, 6, 8, 14], "popul": [2, 6, 8, 9, 11, 12, 14], "being": [2, 8], "enhanc": 2, "friendli": [2, 6], "decai": [2, 8], "creation": [2, 12], "bwlo": 2, "derived_param": [2, 8, 14], "exptc": [2, 14], "par": [2, 8, 14], "dt": [2, 8, 11, 12, 14], "new": [2, 8, 9], "class_nam": [2, 8], "var_init_cod": [2, 8], "extra_global_param": [2, 6, 8, 14], "refer": [2, 3, 8, 9], "read": [2, 8], "repres": [2, 8], "step": [2, 6, 8, 11], "And": [2, 8], "id": [2, 8, 12, 14], "index": [2, 3, 4, 8, 12, 14], "start": [2, 8, 14], "synaps": [2, 3, 6, 8, 9, 11, 12, 14], "id_pr": [2, 8], "id_post": [2, 8], "num_pr": [2, 8], "num_post": [2, 8], "write": [2, 8, 9, 12], "option": [2, 8, 9], "tupl": [2, 8], "resolvedtyp": [2, 8], "callabl": [2, 8], "paramss": [2, 8], "statement": [2, 8], "requir": [2, 4, 6, 8, 9, 14], "want": [2, 6, 8, 9, 11], "redraw": [2, 8], "neg": [2, 8], "ensur": [2, 8], "remain": [2, 8], "causal": [2, 8], "normal_positive_model": [2, 8], "normal_posit": [2, 8], "fly": 2, "row_build_cod": [2, 8], "col_build_cod": [2, 8], "calc_max_row_len_func": [2, 8], "calc_max_col_len_func": [2, 8], "calc_kernel_size_func": [2, 8], "id_post_begin": [2, 8], "first": [2, 6, 8, 12, 14], "addsynaps": [2, 8], "x": [2, 6, 8, 9], "row": [2, 8, 9, 14], "column": [2, 8], "maximum": [2, 8, 12], "length": [2, 8], "matrix": [2, 6, 8], "param_nam": [2, 8], "fix": [2, 8, 12, 14], "replac": [2, 8, 9], "scipi": [2, 8, 12], "stat": [2, 8, 12], "import": [2, 6, 8, 11, 12, 14], "binom": [2, 8], "fixed_number_post": [2, 8], "num": [2, 8, 11, 12, 14], "unsign": [2, 8, 14], "idpost": [2, 8], "ppf": [2, 8, 12], "9999": [2, 8, 12], "our": [2, 8], "paper": [2, 8, 12], "short": [2, 8], "end": [2, 8, 11, 12, 14], "up": [2, 3, 8], "text": [2, 8], "frac": [2, 8], "therefor": [2, 8, 9], "look": [2, 3, 6, 8], "invers": [2, 8], "cummul": [2, 8], "cdf": [2, 8], "chanc": [2, 8], "bound": [2, 8], "correct": [2, 8, 11], "draw": [2, 8], "diagonal_build_cod": [2, 8], "diagon": [2, 8, 9], "independ": [2, 8], "id_diag": [2, 8], "id_kern_0": [2, 8], "id_kern_1": [2, 8], "id_kern_n": [2, 8], "dimension": [2, 8], "for_each_synaps": [2, 8, 9], "construct": [2, 3, 6, 8], "loop": [2, 6, 8, 9, 12, 14], "incom": [2, 8], "insid": [2, 8], "convolv": [2, 8], "kern_dim": [2, 8], "squar": [2, 8, 14], "pop_dim": [2, 8], "simple_conv2d_model": [2, 8], "pynn": [2, 8], "simple_conv2d": [2, 8], "kern_siz": [2, 8], "kernrow": [2, 8], "kerncol": [2, 8], "prerow": [2, 8], "precol": [2, 8], "If": [2, 3, 4, 6, 8, 14], "haven": [2, 4, 8], "gone": [2, 8], "off": [2, 8], "edg": [2, 8], "postrow": [2, 8], "postcol": [2, 8], "postind": [2, 8], "express": [2, 8, 14], "extend": [2, 6, 8], "abov": [2, 4, 6, 8, 9], "sim_cod": [2, 8, 9, 11, 14], "threshold_condition_cod": [2, 8, 11, 14], "reset_cod": [2, 8, 11, 14], "additional_input_var": [2, 8, 14], "auto_refractory_requir": [2, 8], "fals": [2, 8, 11, 12, 14], "isyn": [2, 8, 11, 14], "total": [2, 5, 8, 12, 13], "modifi": [2, 4, 8, 9], "threshold": [2, 8], "condit": [2, 8], "test": [2, 8, 11], "emit": [2, 6, 8], "list": [2, 8, 12], "bool": [2, 8], "doe": [2, 8, 9], "auto": [2, 8, 11], "refractori": [2, 8], "logic": [2, 8], "leaki": [2, 8], "integr": [2, 8, 11, 14], "dv": [2, 8], "i_": [2, 8], "rm": [2, 8], "solv": [2, 8], "euler": [2, 8], "method": [2, 6, 8], "leaky_integrator_model": [2, 8], "leaky_integr": [2, 8], "read_writ": [2, 8], "receiv": [2, 8], "linear": [2, 8], "sum": [2, 8, 11, 14], "come": [2, 8], "linearli": [2, 8], "product": [2, 8], "isyn2": [2, 8], "event": [2, 6, 8], "driven": [2, 8], "pre_": [2, 8], "post_": [2, 8], "pre_neuron_var_ref": [2, 8, 14], "post_neuron_var_ref": [2, 8, 14], "pre_spike_syn_cod": [2, 8, 11, 14], "pre_event_syn_cod": [2, 8], "post_event_syn_cod": [2, 8], "post_spike_syn_cod": [2, 8, 11], "synapse_dynamics_cod": [2, 8, 14], "pre_spike_cod": [2, 8, 14], "post_spike_cod": [2, 8], "pre_dynamics_cod": [2, 8, 14], "post_dynamics_cod": [2, 8, 14], "assumpt": [2, 8], "addtopost": [2, 8, 14], "inc": [2, 8], "amount": [2, 8], "dendrit": [2, 8, 12], "insert": [2, 8], "addtopostdelai": [2, 8], "again": [2, 4, 8], "heterogen": [2, 8], "weightupdatemodel": [2, 8], "staticpulsedendriticdelai": [2, 8, 12], "simpl": [2, 3, 8], "max_dendritic_delay_timestep": [2, 8, 12], "properti": [2, 6, 8], "effect": [2, 8], "revers": [2, 8, 12], "direct": [2, 8], "addtopr": [2, 8], "v_post": [2, 8], "_outgoing_": [2, 8], "pre_target_var": [2, 8], "unlik": [2, 8], "action": [2, 8, 11, 12, 14], "modul": [2, 3, 7], "directli": [2, 6, 8, 9], "indic": [2, 6, 8, 12, 14], "varaccessmod": [2, 8], "pre_event_threshold_condition_cod": [2, 8], "post_event_threshold_condition_cod": [2, 8], "stdp": [2, 8, 11], "rule": [2, 8, 11, 14], "nearest": [2, 8], "neighbour": [2, 8], "depend": [2, 8], "delta": [2, 8], "w_": [2, 8], "ij": [2, 8], "begin": [2, 8], "case": [2, 8, 9], "a_": [2, 8], "left": [2, 8], "tau_": [2, 8], "right": [2, 8], "leq0": [2, 8], "manner": [2, 8], "stdp_additive_model": [2, 8], "stdp_addit": [2, 8], "tauplu": [2, 8], "tauminu": [2, 8], "aplu": [2, 8], "aminu": [2, 8], "wmin": [2, 8, 11, 14], "wmax": [2, 8, 11, 14], "st_post": [2, 8, 11], "newweight": [2, 8, 11], "st_pre": [2, 8, 11], "cost": [2, 8], "tend": [2, 8], "grow": [2, 8], "o": [2, 8], "basi": [2, 8], "good": [2, 8], "idea": [2, 8], "pre_var_name_typ": [2, 8], "post_var_name_typ": [2, 8], "_trace_": [2, 8], "stdp_additive_2_model": [2, 8], "genn_model": [2, 8], "create_custom_weight_update_class": [2, 8], "stdp_additive_2": [2, 8], "pretrac": [2, 8], "posttrac": [2, 8], "tauplusdecai": [2, 8], "tauminusdecai": [2, 8], "previous": [2, 6, 8, 9], "intern": [2, 8], "continu": [2, 8], "od": [2, 8], "computation": [2, 8], "costli": [2, 8], "becaus": [2, 6, 8], "larg": [2, 8], "discuss": [2, 6, 8], "rate": [2, 8, 12, 14], "contin": [2, 8], "multipli": [2, 8, 12], "definit": [2, 8], "v_pre": [2, 8, 9], "evalu": [2, 8, 9], "involv": [2, 8], "respect": [2, 8, 9], "voltag": [2, 8], "greater": [2, 8], "whenev": [2, 8], "true": [2, 6, 8, 11, 12, 14], "pre_event_cod": [2, 8], "equat": [2, 8], "injectcurr": [2, 8, 9, 11], "goe": [2, 8], "post_target_var": [2, 8, 14], "injection_cod": [2, 8, 11], "target_var": [2, 8], "uniform_noise_model": [2, 8], "uniform_nois": [2, 8], "magnitud": [2, 8, 11], "demand": 2, "update_cod": [2, 8, 14], "extra_global_param_ref": [2, 8], "launch": [2, 6, 8], "reset_model": [2, 8], "read_onli": [2, 8], "sort": [2, 8, 9, 14], "read_only_dupl": [2, 8], "mode": [2, 8], "reduct": [2, 8], "reduce_batch_sum": [2, 8], "reduce_batch_max": [2, 8], "reduce_model": [2, 8], "gradient_batch_reduc": [2, 8], "reducedgradi": [2, 8], "reduce_sum": [2, 8], "reduce_max": [2, 8], "similarli": [2, 6, 8], "reduce_neuron_sum": [2, 8], "reduce_neuron_max": [2, 8], "neuron_reduc": [2, 8], "row_update_cod": [2, 8], "host_update_cod": [2, 8], "design": [2, 8], "issu": [2, 8], "regard": [2, 8], "accompani": [2, 8], "push": [2, 8, 11], "pull": [2, 8], "illustr": [2, 8], "abil": [2, 8], "remov": [2, 8], "remove_diagonal_model": [2, 8], "remove_diagon": [2, 8], "remove_synaps": [2, 8], "break": [2, 8, 14], "back": [2, 8], "add_diagonal_model": [2, 8], "add_diagon": [2, 8], "add_synaps": [2, 8], "lot": [2, 8], "_might_": [2, 8], "detect": [2, 8], "shuffl": [2, 8], "around": [2, 8], "accordingli": [2, 8], "fine": [2, 8], "know": [2, 8], "hook": [2, 8], "abl": [2, 8], "common": [2, 6, 8], "parallel": [2, 8], "determin": [2, 4, 8, 11, 12], "element": [2, 6, 8], "get": [2, 4, 8, 12, 14], "befor": [2, 6, 8], "pushpostindtodevic": [2, 8], "softwar": 3, "packag": [3, 4, 7], "nvidia": [3, 4], "api": [3, 6], "neuron": [3, 6, 8, 9, 11, 12, 14], "genncod": 3, "under": [3, 12, 14], "find": 3, "contact": 3, "project": [3, 11, 12], "develop": [3, 4, 10, 12, 14], "upgrad": 3, "4": [3, 8, 12], "custom": [3, 8, 11, 14], "bibliographi": 3, "maintain": [3, 9], "dr": 3, "jame": 3, "prof": 3, "thoma": 3, "partial": [3, 14], "epsrc": 3, "grant": 3, "ep": 3, "v052241": 3, "unlock": 3, "research": 3, "p006094": 3, "board": 3, "j019690": 3, "green": 3, "search": 3, "page": 3, "futur": 4, "plan": 4, "binari": 4, "conda": 4, "now": [4, 9], "sourc": [4, 8, 9, 10, 11, 12, 14], "compil": [4, 8, 9], "alreadi": [4, 8], "window": [4, 8], "visual": 4, "studio": 4, "2019": 4, "microsoft": 4, "commun": 4, "edit": 4, "download": [4, 10, 11, 12, 14], "www": 4, "visualstudio": 4, "vs": 4, "aspx": 4, "desktop": 4, "linux": 4, "gnu": 4, "collect": [4, 6], "gcc": 4, "7": [4, 8, 12], "obtain": [4, 8], "repositori": 4, "ubuntu": 4, "sudo": 4, "apt": 4, "html": [4, 8], "fresh": 4, "toolkit": 4, "sure": 4, "pick": [4, 8], "compat": [4, 9], "latest": 4, "necessarili": 4, "cuda_path": 4, "environ": 4, "variabl": [4, 8, 9, 11, 14], "against": 4, "choos": 4, "verifi": 4, "echo": 4, "command": 4, "prompt": 4, "mac": 4, "bash": 4, "export": 4, "usr": 4, "locat": [4, 6, 8], "persist": 4, "login": 4, "script": 4, "profil": [4, 8, 12, 14], "bashrc": 4, "releas": [4, 8], "extract": [4, 6], "home": 4, "directori": [4, 8], "clone": 4, "git": 4, "github": 4, "team": 4, "libffi": 4, "dev": 4, "pybind11": 4, "psutil": 4, "numpi": [4, 11, 12, 14], "wish": [4, 6], "most": [4, 8, 9], "intend": [4, 8], "yourself": 4, "least": [], "8": [8, 12], "navig": [], "library_directori": [], "pwd": [], "genn_wrapp": [], "2015": [], "extens": [], "earlier": [], "11": 8, "instruct": [], "anaconda": [], "wa": [6, 9], "_should_": [], "suitabl": [], "let": 9, "suce": [], "correctli": 8, "x64": [], "nativ": [], "tool": [], "chosen": [6, 12], "menu": [], "folder": [], "_activate_": [], "bat": [], "me": [], "anaconda3": [], "swig": [], "dll": [], "msbuild": [], "sln": [], "release_dl": [], "don": [], "fail": [], "still": [], "newli": [], "y": 8, "lib": [], "00": [5, 13], "000": [5, 13], "file": [5, 8, 13, 14], "galleri": [5, 10, 11, 12, 14], "mem": [5, 13], "mb": [5, 13], "mnist": [5, 10, 13], "classif": [5, 10, 13], "insect": [5, 10, 13], "inspir": [5, 10, 13], "mushroom": [5, 10, 13], "bodi": [5, 10, 13], "userproject": [5, 8, 13], "mnist_mb_classifi": [5, 11, 13], "py": [5, 11, 12, 13, 14], "potjans_microcircuit": [5, 12, 13], "superspike_demo": [5, 13, 14], "load": [6, 8, 11, 12, 14], "lazi": 6, "hasn": 6, "almost": [6, 9], "instantan": [6, 8], "error": [6, 8, 14], "report": 6, "simplest": 6, "100": [6, 11, 12], "step_tim": [6, 8, 11, 12, 14], "ms": [6, 8, 12, 14], "asynchron": [6, 12], "synchronis": 6, "natur": [6, 8], "ineffici": [6, 8, 12], "dedic": 6, "transfer": 6, "spike_recording_en": [6, 8, 11, 12, 14], "spike_event_recording_en": [6, 8], "runtim": [6, 7, 9], "num_recording_timestep": [6, 8, 11, 12, 14], "pull_recording_buffers_from_devic": [6, 8, 11, 12, 14], "spike_recording_data": [6, 8, 11, 12, 14], "spike_event_recording_data": 6, "spike_tim": [6, 11, 12], "spike_id": [6, 12], "real": 6, "interact": 6, "encapsul": 6, "model_preprocessor": [6, 7], "variablebas": [6, 8], "object": [6, 8], "own": 6, "wherea": [6, 9], "content": 6, "arraybas": [6, 8], "push_to_devic": [6, 8, 11, 14], "pull_from_devic": [6, 8, 11, 14], "noth": 6, "recommend": 6, "leav": 6, "work": [6, 8], "transparantli": 6, "current_valu": [6, 8], "npy": [6, 11], "transform": [6, 8, 14], "format": [6, 8, 14], "matric": 6, "re": [6, 12, 14], "order": [6, 8, 9, 12, 14], "current_view": [6, 8], "behav": 6, "extraglobalparamet": [6, 8], "hold": [6, 8], "updat": [6, 8, 9, 12, 14], "psm_extra_global_param": 6, "set_init_valu": [6, 8, 14], "just": [6, 9], "set_param_dynam": [6, 8, 14], "groupmixin": [6, 8], "set_dynamic_param_valu": [6, 8, 14], "increas": [6, 8], "parameterm": 6, "submodul": 7, "cuda_backend": 7, "genn_group": 7, "postsynaptic_model": 7, "single_threaded_cpu_backend": 7, "pybind11_object": 8, "currentsourcemixin": 8, "get_var_loc": 8, "self": 8, "_genn": 8, "arg0": 8, "set_var_loc": 8, "arg1": 8, "var_initialis": 8, "customconnectivityupd": 8, "customconnectivityupdatemixin": 8, "get_post_var_loc": 8, "get_pre_var_loc": 8, "post_var_initialis": 8, "post_var_refer": 8, "pre_var_initialis": 8, "pre_var_refer": 8, "set_post_var_loc": 8, "ignor": 8, "space": 8, "set_pre_var_loc": 8, "synapse_group": 8, "update_group_nam": 8, "var_refer": 8, "customupdatebas": 8, "customupdatemixin": 8, "ax": [8, 12, 14], "subtract": 8, "ie": 8, "varaccessdim": 8, "axi": [8, 11, 14], "written": 8, "whatev": [8, 9], "read_only_shar": 8, "asid": 8, "read_only_shared_neuron": 8, "33": 8, "84": 8, "76": 8, "52": 8, "44": 8, "customupdatewumixin": 8, "model_nam": 8, "time_precis": 8, "genn_log_level": 8, "plogsever": 8, "warn": 8, "code_gen_log_level": 8, "transpiler_log_level": 8, "runtime_log_level": 8, "backend_log_level": 8, "preference_kwarg": 8, "modelspecintern": 8, "interfac": 8, "best": 8, "level": 8, "transpil": 8, "backend_nam": 8, "path_to_model": 8, "always_rebuild": 8, "never_rebuild": 8, "path": 8, "rebuilt": 8, "even": 8, "doesn": [8, 12], "appear": 8, "never": 8, "ever": 8, "prevent": 8, "overwrit": 8, "custom_upd": [8, 14], "get_custom_update_tim": [8, 14], "second": [8, 12, 14], "spent": 8, "kernel_tim": 8, "get_custom_update_transpose_tim": [8, 14], "init_sparse_tim": [8, 12, 14], "init_tim": [8, 12, 14], "record": [8, 11, 12, 14], "neuron_update_tim": [8, 12, 14], "postsynaptic_update_tim": 8, "presynaptic_update_tim": [8, 12, 14], "buffer": 8, "synapse_dynamics_tim": [8, 14], "unload": 8, "free": 8, "neurongroupmixin": 8, "num_delay_slot": 8, "prev_spike_time_loc": 8, "prev_spike_time_requir": 8, "recording_zero_copy_en": 8, "spike_time_loc": 8, "spike_time_requir": 8, "parallelismhint": 8, "word_packed_bitmask": 8, "fatal": 8, "info": 8, "verbos": 8, "synapsegroupmixin": 8, "axonal_delay_step": [8, 9], "back_prop_delay_step": 8, "dendritic_delay_loc": 8, "get_ps_var_loc": 8, "get_wu_post_var_loc": 8, "get_wu_pre_var_loc": 8, "get_wu_var_loc": 8, "kernel_s": 8, "max_connect": 8, "max_source_connect": 8, "narrow_sparse_ind_en": 8, "num_threads_per_spik": [8, 12], "output_loc": 8, "parallelism_hint": 8, "ps_initialis": 8, "set_ps_param_dynam": 8, "set_ps_var_loc": 8, "set_wu_param_dynam": 8, "set_wu_post_var_loc": 8, "set_wu_pre_var_loc": 8, "set_wu_var_loc": 8, "sparse_connectivity_initialis": 8, "sparse_connectivity_loc": 8, "toeplitz_connectivity_initialis": 8, "wu_initialis": 8, "synapsematrixconnect": 8, "flag": 8, "differnet": 8, "16": 8, "66": 8, "129": 8, "136": 8, "264": 8, "68": 8, "272": 8, "synapsematrixweight": 8, "256": [8, 14], "128": 8, "97": 8, "varaccessmodeattribut": 8, "attribut": 8, "summat": 8, "zero_copi": 8, "arg": [8, 11, 12, 14], "kwarg": 8, "create_post_var_ref": 8, "create_pre_var_ref": 8, "create_psm_egp_ref": 8, "create_wu_egp_ref": 8, "var_nam": [8, 11], "transpose_sg": 8, "transpose_var_nam": 8, "get_var_access_dim": 8, "blocksizeselect": 8, "block": 8, "occup": 8, "blocksiz": 8, "deviceselect": 8, "most_memori": 8, "preferencesbas": 8, "block_size_select_method": 8, "constant_cache_overhead": 8, "four": 8, "header": [8, 12, 14], "neuronupd": 8, "synapseupd": 8, "runner": 8, "take": [8, 9], "72": 8, "byte": 8, "lookup": 8, "tabl": 8, "curand": 8, "applic": 8, "device_select_method": 8, "enable_nccl_reduct": 8, "correspond": 8, "nccl": 8, "generate_line_info": 8, "line": 8, "purpos": 8, "manual_block_s": 8, "show_ptx_info": 8, "ptx": 8, "assembl": 8, "inform": 8, "displai": [8, 14], "dure": 8, "dc": [8, 12], "It": [8, 12, 14], "amp": 8, "amplitud": 8, "na": [8, 11, 12], "noisi": 8, "poissonexp": [8, 12], "equival": 8, "poisson": [8, 14], "tausyn": [8, 12], "fire": [8, 11, 12], "hz": [8, 12], "minim": [8, 11], "get_var_valu": 8, "parent": 8, "pull_extra_global_param_from_devic": 8, "egp_nam": 8, "pull_var_from_devic": 8, "given": [8, 12], "push_extra_global_param_to_devic": 8, "push_var_to_devic": 8, "get_sparse_post_ind": [8, 11], "get_sparse_pre_ind": [8, 11], "post_spike_event_recording_data": 8, "pre_spike_event_recording_data": 8, "pull_connectivity_from_devic": [8, 11], "pull_in_syn_from_devic": 8, "pull_psm_extra_global_param_from_devic": 8, "wrapper": 8, "push_connectivity_to_devic": 8, "push_in_syn_to_devic": 8, "push_psm_extra_global_param_to_devic": 8, "set_sparse_connect": [8, 11], "pre_indic": 8, "post_indic": 8, "rag": 8, "weight_update_var_s": 8, "convert": [8, 11, 14], "channel": 8, "rang": [8, 11, 14], "fixednumberpostwithreplac": 8, "random": [8, 14], "discret": 8, "uniform": 8, "ascend": 8, "1st": 8, "statist": 8, "beta": [8, 14], "npost": 8, "next": 8, "smallest": 8, "special": 8, "fixednumberprewithreplac": [8, 11], "fixednumbertotalwithreplac": [8, 12], "stage": 8, "multinomi": 8, "parallelis": 8, "exist": 8, "bernoulli": 8, "repeatedli": 8, "geometr": 8, "trial": [8, 14], "success": 8, "gap": 8, "devroy": 8, "1986": 8, "invert": 8, "fixedprobabilitynoautaps": 8, "autaps": 8, "recurr": 8, "br": 8, "inneffici": 8, "gemetr": 8, "onetoon": 8, "uninitialis": 8, "mark": 8, "avgpoolconv2d": 8, "averag": [8, 11, 14], "pool": 8, "intialis": 8, "seldom": 8, "initvarsnippet": 8, "implicit": 8, "constructor": 8, "unit": [8, 14], "distanc": 8, "normalclip": [8, 12, 14], "resampl": 8, "out": 8, "my": 8, "thgenn": 8, "minimum": 8, "normalclippeddelai": [8, 12], "variable_typ": 8, "view": [8, 11, 14], "set_arrai": 8, "view_shap": 8, "variable_nam": 8, "init_valu": 8, "about": 8, "iter": 8, "set_valu": 8, "synapsevari": 8, "delay_group": 8, "valinit_valuesu": 8, "cite": 8, "izhikevich2003simpl": 8, "eqnarrai": 8, "04": 8, "140": 8, "du": 8, "bv": 8, "extern": [8, 12], "increment": 8, "mv": 8, "particular": 8, "popular": 8, "though": 8, "due": 8, "strictli": 8, "inconsist": 8, "membran": [8, 14], "potenti": [8, 14], "recoveri": 8, "sensit": 8, "izhikevichvari": 8, "neuronmodel": 8, "lif": [8, 11, 12], "vrest": [8, 11, 12, 14], "unless": 8, "randomli": 8, "vspike": 8, "spiketim": [8, 14], "trefract": 8, "last": 8, "period": 8, "tspike": 8, "durat": [8, 12], "rest": 8, "entri": [8, 12, 14], "That": 8, "equal": 8, "undefin": 8, "firingprob": 8, "cdot": 8, "pattern": 8, "leq": 8, "approxim": 8, "math": 8, "relev": 8, "especi": 8, "quit": 8, "small": 8, "moder": 8, "worth": 8, "becom": [8, 9], "poor": 8, "poissonnew": 8, "accord": 8, "timesteptospik": 8, "rulkovmap": 8, "rulkov": 8, "map": 8, "rulkov2002": 8, "nowotny2005self": 8, "ll": 8, "v_": 8, "big": 8, "otherwis": [8, 14], "prev": 8, "60mv": 8, "shift": 8, "excit": 8, "origin": [8, 12], "468": 8, "roughli": 8, "resist": 8, "regul": 8, "omega": 8, "spikesourc": 8, "empti": 8, "spikegeneratorgroup": 8, "brian": 8, "spikesourcearrai": [8, 14], "startspik": [8, 14], "endspik": [8, 14], "globel": 8, "traubmil": 8, "hodgkin": 8, "huxlei": 8, "traub": 8, "mile": 8, "algorithm": 8, "taken": 8, "traub1991": 8, "i_k": 8, "leak": 8, "i_m": 8, "i_i": 8, "g_": 8, "m_i": 8, "h_i": 8, "v_i": 8, "e_": 8, "k": 8, "n_i": 8, "dy": 8, "alpha_i": 8, "beta_i": 8, "y_i": 8, "h": [8, 11, 12, 14], "alpha_n": 8, "032": 8, "50": [8, 11, 12, 14], "beta_n": 8, "55": 8, "40": 8, "alpha_m": 8, "beta_m": 8, "28": 8, "25": [8, 12], "alpha_h": 8, "48": 8, "18": 8, "beta_h": 8, "143": 8, "nf": 8, "02672": 8, "mu": 8, "63": 8, "563": 8, "15": [8, 11], "43": 8, "95": 8, "gna": 8, "mohm": 8, "cm": 8, "ena": 8, "equi": 8, "gk": 8, "ek": 8, "gl": 8, "el": 8, "capac": [8, 11], "densiti": 8, "muf": 8, "ordinari": 8, "differenti": 8, "ldt": 8, "004": 8, "variant": [8, 9], "IF": [8, 11], "check": 8, "singular": 8, "hit": 8, "l": 8, "hospit": 8, "traubmilesalt": 8, "workaround": 8, "avoid": 8, "munimum": 8, "traubmilesfast": 8, "fast": 8, "inner": 8, "There": 8, "traubmilesnstep": 8, "deltacurr": [8, 11, 14], "Is": 8, "expdecai": 8, "expf": 8, "piecewisestdp": 8, "finit": 8, "transmiss": 8, "piecewis": 8, "imag": [8, 11], "learn1synapse_explain_html": 8, "png": 8, "latex": 8, "learn1synapse_explain": 8, "width": 8, "10cm": 8, "curv": 8, "raw": 8, "graw": 8, "filter": [8, 14], "sugmoid": 8, "impli": 8, "unpair": 8, "incur": 8, "henc": 8, "stxx": 8, "xx": [8, 9], "somewhat": [8, 12], "arbitrarili": 8, "subject": 8, "sigmoid": 8, "revert": 8, "map_classol": 8, "cc": 8, "mbody1": 8, "neuronn": 8, "gkcdn": 8, "scalar_min": 8, "cnt": 8, "fprintf": 8, "stdout": 8, "too": 8, "low": [8, 12], "below": [8, 10], "tmp": 8, "mykcdn_p": 8, "grawkcdn": 8, "cerr": 8, "endl": 8, "lead": 8, "infin": 8, "nomin": 8, "extrem": 8, "act": 8, "g_0": 8, "t_": 8, "compar": 8, "figur": [8, 12], "tlrn": 8, "tchng": 8, "tdecai": 8, "strength": 8, "tpunish10": 8, "suppress": 8, "respons": 8, "tpunish01": 8, "gmax": 8, "maxim": 8, "achiev": [8, 9], "gmid": 8, "midpoint": 8, "gslope": 8, "slope": 8, "taushift": 8, "gsyn0": 8, "staticgrad": 8, "grade": 8, "gradual": 8, "gsyn": 8, "larger": 8, "epr": 8, "vslope": 8, "_pre": 8, "_post": 8, "staticpuls": [8, 9, 11], "coupl": 8, "sim": 8, "aim": 9, "backward": 9, "strive": 9, "underli": 9, "pars": 9, "subset": 9, "old": 9, "necessari": [9, 12], "streamlin": 9, "area": 9, "were": 9, "apply_input_cod": 9, "decay_cod": 9, "unnecessarili": 9, "complex": 9, "wasn": 9, "obviou": 9, "cumbersom": 9, "wors": 9, "axon": 9, "realli": 9, "ugli": 9, "confus": 9, "outsid": 9, "reus": 9, "globalg": 9, "individualg": 9, "Then": 9, "ve": 9, "renam": 9, "chose": 9, "unusu": 9, "creatabl": 9, "pointer": 9, "arbitrari": 9, "latter": 9, "handl": 9, "_implicit_": 9, "_explicit_": 9, "userproject_python": 10, "zip": [10, 14], "jupyt": [10, 11, 12, 14], "notebook": [10, 11, 12, 14], "userproject_jupyt": 10, "sphinx": [10, 11, 12, 14], "digit": 11, "usag": [11, 12, 14], "plot": [11, 12, 14], "argpars": [11, 12, 14], "argumentpars": [11, 12, 14], "tqdm": 11, "factor": [11, 12], "normalis": 11, "pixel": 11, "input_scal": 11, "stimul": 11, "mbon": 11, "mbon_stimulus_curr": 11, "match": [11, 12], "num_pn": 11, "784": 11, "kenyon": 11, "num_kc": 11, "20000": 11, "num_mbon": 11, "present_time_m": 11, "lif_param": [11, 12], "taum": [11, 12], "60": [11, 14], "vreset": [11, 12], "vthresh": [11, 12, 14], "ioffset": [11, 12], "taurefrac": [11, 12, 14], "pn": 11, "pn_param": 11, "pn_kc_weight": 11, "pn_kc_tau_syn": 11, "pn_kc_fan_in": 11, "kc": 11, "ggn": 11, "inhibit": 11, "200": [11, 14], "ggn_param": 11, "kc_mbon_tau_syn": 11, "kc_mbon_param": 11, "rho": 11, "01": 11, "eta": 11, "00002": 11, "cs_model": 11, "if_model": 11, "symmetr": 11, "symmetric_stdp": 11, "cli": [11, 14], "def": [11, 12, 14], "get_pars": [11, 12, 14], "parser": [11, 12, 14], "add_argu": [11, 12, 14], "store_tru": [11, 12, 14], "help": [11, 12, 14], "__name__": [11, 12, 14], "__main__": [11, 12, 14], "parse_arg": [11, 12, 14], "reshap": [11, 14], "test_imag": 11, "els": [11, 12, 14], "train_imag": 11, "astyp": 11, "float32": [11, 12], "newaxi": 11, "label": 11, "test_label": 11, "train_label": 11, "mnist_mb": 11, "lif_init": [11, 12], "refractim": [11, 12, 14], "if_init": 11, "turn": [11, 14], "pn_input": 11, "mbon_input": 11, "pn_kc_connect": 11, "pn_kc": 11, "pn_kc_ind": 11, "kc_ggn": 11, "ggn_kc": 11, "kc_mbon_weight_upd": 11, "kc_mbon_g": 11, "kc_mbon": 11, "present_timestep": 11, "reset_spike_tim": 11, "finfo": [11, 12], "reset_out_post": 11, "out_post": 11, "reset_neuron": 11, "var_init": 11, "var_val": 11, "item": [11, 12], "num_correct": 11, "count": [11, 14], "mbon_spike_tim": 11, "mbon_spike_id": 11, "len": [11, 12, 14], "argmin": 11, "print": [11, 12, 14], "weigh": 11, "kc_mbon_g_view": 11, "vstack": [11, 14], "plot_weight_distribut": 11, "matplotlib": [11, 12, 14], "pyplot": [11, 12, 14], "plt": [11, 12, 14], "fig": [11, 12, 14], "subplot": [11, 12, 14], "figsiz": 11, "hist": 11, "bin": 11, "axvlin": 11, "linestyl": 11, "set_xlabel": [11, 12, 14], "set_ylabel": [11, 12, 14], "show": [11, 12, 14], "ipynb": [11, 12, 14], "reimplement": [12, 14], "tobia": 12, "marku": 12, "spontan": 12, "irregular": 12, "agreement": 12, "vivo": 12, "awak": 12, "anim": 12, "excitatori": 12, "neuron_scal": 12, "connectivity_scal": 12, "1000": [12, 14], "norm": 12, "perf_count": 12, "layer_nam": 12, "23": 12, "population_nam": 12, "dt_m": 12, "background": 12, "background_r": 12, "rel": 12, "inhibitori": 12, "except": 12, "l4e": 12, "l2": 12, "3e": 12, "mean_w": 12, "87": 12, "8e": 12, "external_w": 12, "801": 12, "paragraph": 12, "parameter": 12, "caption": 12, "supplementari": 12, "layer_23_4_w": 12, "rel_w": 12, "mention": 12, "layer_23_4_relw": 12, "05": [12, 14], "20683": 12, "5834": 12, "21915": 12, "5479": 12, "4850": 12, "1065": 12, "14395": 12, "2948": 12, "connection_probabilti": 12, "23e": 12, "1009": 12, "23i": 12, "1689": 12, "4e": 12, "0437": 12, "4i": 12, "0818": 12, "5e": 12, "0323": 12, "5i": 12, "6e": 12, "0076": 12, "6i": 12, "1346": 12, "1371": 12, "0316": 12, "0515": 12, "0755": 12, "0042": 12, "0077": 12, "0059": 12, "0497": 12, "135": 12, "0067": 12, "0003": 12, "0453": 12, "0691": 12, "0029": 12, "0794": 12, "1597": 12, "0033": 12, "1057": 12, "1004": 12, "0622": 12, "0505": 12, "0057": 12, "0831": 12, "3726": 12, "0204": 12, "0548": 12, "0269": 12, "0257": 12, "0022": 12, "06": 12, "3158": 12, "0086": 12, "0156": 12, "0066": 12, "0211": 12, "0166": 12, "0572": 12, "0197": 12, "0396": 12, "2252": 12, "0364": 12, "001": [12, 14], "0034": 12, "0005": 12, "0277": 12, "0658": 12, "1443": 12, "degre": 12, "num_external_input": 12, "1600": 12, "1500": 12, "2100": 12, "1900": 12, "2000": 12, "2900": 12, "realiz": 12, "mean_firing_r": 12, "971": 12, "868": 12, "746": 12, "396": 12, "142": 12, "9": [12, 14], "078": 12, "991": 12, "523": 12, "mean_delai": 12, "75": 12, "delay_sd": 12, "375": 12, "helper": [12, 14], "get_scaled_num_neuron": 12, "get_full_num_input": 12, "src_layer": 12, "trg_layer": 12, "trg_pop": 12, "num_src": 12, "num_trg": 12, "connection_prob": 12, "get_mean_weight": 12, "get_scaled_num_connect": 12, "num_input": [12, 14], "assert": 12, "get_full_mean_input_curr": 12, "mean_input_curr": 12, "fuse_postsynaptic_model": 12, "default_narrow_sparse_ind_en": 12, "timing_en": [12, 14], "kernel_profil": [12, 14], "default_var_loc": 12, "default_sparse_connectivity_loc": 12, "58": 12, "poisson_init": 12, "exp_curr_init": 12, "quantil": 12, "normal_quantile_cdf": 12, "max_delai": 12, "fm": 12, "slot": 12, "seem": 12, "aggress": 12, "merg": 12, "max_dendritic_delay_slot": 12, "total_neuron": 12, "neuron_popul": 12, "ext_input_r": 12, "ext_weight": 12, "ext_input_curr": 12, "poisson_param": 12, "pop_siz": 12, "neuron_pop": 12, "_poisson": 12, "tpopul": 12, "total_synaps": 12, "num_sub_row": 12, "procedural_connect": 12, "trg_name": 12, "src_name": 12, "mean_weight": 12, "weight_sd": 12, "num_connect": 12, "tconnect": 12, "numconnect": 12, "meanweight": 12, "weightsd": 12, "meandelai": 12, "delaysd": 12, "connector": 12, "connect_param": 12, "d_dist": 12, "synapse_nam": 12, "hack": 12, "cast": 12, "w_dist": 12, "static_synapse_init": 12, "syn_pop": 12, "span": 12, "duration_timestep": 12, "ten_percent_timestep": 12, "sim_start_tim": 12, "advanc": 12, "sim_end_tim": 12, "tsimul": 12, "tinit": 12, "tspars": 12, "tneuron": 12, "tsynaps": 12, "save_data": [12, 14], "csv": [12, 14], "savetxt": [12, 14], "_spike": 12, "column_stack": [12, 14], "delimit": [12, 14], "fmt": [12, 14], "yuck": 12, "ordered_neuron_popul": 12, "start_id": 12, "bar_i": 12, "actor": 12, "scatter": [12, 14], "edgecolor": [12, 14], "bar": 12, "colour": 12, "barh": 12, "align": 12, "center": 12, "color": 12, "get_facecolor": 12, "ecolor": 12, "black": 12, "po": 12, "firingr": 12, "set_ytick": 12, "set_yticklabel": 12, "friedemann": 14, "surya": 14, "radcliff": 14, "camera": 14, "oxford": 14, "record_tri": 14, "target_fil": 14, "num_trial": 14, "filenam": 14, "ra": 14, "600": 14, "timestep_m": 14, "num_output": 14, "num_hidden": 14, "tau_rise_m": 14, "tau_decay_m": 14, "tau_rms_m": 14, "30000": 14, "tau_avg_err_m": 14, "10000": 14, "r0": 14, "epsilon": 14, "1e": 14, "tau_decay_": 14, "tau_rise_": 14, "tau_avg_err_": 14, "scale_tr_err_flt": 14, "auryn": 14, "volt": 14, "1000x": 14, "w_min": 14, "w_max": 14, "w0": 14, "experi": 14, "input_freq_hz": 14, "update_time_m": 14, "500": 14, "trial_m": 14, "1890": 14, "update_timestep": 14, "trial_timestep": 14, "calc_t_peak": 14, "tau_ris": 14, "tau_decai": 14, "write_spike_fil": 14, "r_max_prop_model": 14, "r_max_prop": 14, "updatetim": 14, "taurm": 14, "upsilon": 14, "updatetimestep": 14, "exprm": 14, "superspike_model": 14, "tauris": 14, "taudecai": 14, "z": 14, "ztilda": 14, "sigmaprim": 14, "errtilda": 14, "trace": 14, "oneplushi": 14, "elig": 14, "feedback_model": 14, "feedback": 14, "hidden_neuron_model": 14, "hidden": 14, "taumem": 14, "isynfeedback": 14, "rmembran": 14, "output_neuron_model": 14, "tauavgerr": 14, "errris": 14, "avgsqrerr": 14, "errdecai": 14, "normfactor": 14, "trisemult": 14, "tdecaymult": 14, "tpeak": 14, "mulavgerr": 14, "spred": 14, "sreal": 14, "mismatch": 14, "temp": 14, "narg": 14, "target_spik": 14, "loadtxt": 14, "dtype": 14, "neuron_id": 14, "millisecond": 14, "target_neuron_end_tim": 14, "cumsum": 14, "bincount": 14, "minlength": 14, "target_neuron_start_tim": 14, "concaten": 14, "frozen": 14, "input_isi_m": 14, "input_spike_tim": 14, "vector": 14, "reach": 14, "stack": 14, "input_spikes_per_neuron": 14, "togeth": 14, "input_spik": 14, "input_neuron_end_tim": 14, "input_neuron_start_tim": 14, "input_init_var": 14, "hidden_param": 14, "hidden_init_var": 14, "output_param": 14, "output_init_var": 14, "superspike_param": 14, "superspike_pre_init_var": 14, "superspike_post_init_var": 14, "input_hidden_weight_dist_param": 14, "input_hidden_init_var": 14, "hidden_output_weight_dist_param": 14, "hidden_output_init_var": 14, "r_max_prop_param": 14, "descript": 14, "generatelineinfo": 14, "any_record": 14, "input_hidden": 14, "inputhidden": 14, "hidden_output": 14, "hiddenoutput": 14, "output_hidden": 14, "outputhidden": 14, "input_hidden_transpos": 14, "calculatetranspos": 14, "input_hidden_optimiser_var_ref": 14, "input_hidden_optimis": 14, "gradientlearn": 14, "hidden_output_optimiser_var_ref": 14, "hidden_output_optimis": 14, "output_avg_sqr_err_var": 14, "current_r0": 14, "hidden_spik": 14, "output_spik": 14, "perid": 14, "time_": 14, "mean_error": 14, "0e": 14, "upload": 14, "repeat": 14, "input_spikes_": 14, "hidden_spikes_": 14, "output_spikes_": 14, "append": 14, "sharex": 14, "col": 14, "sharei": 14, "enumer": 14, "start_time_": 14, "890": 14, "instal": 3, "build_ext": 4}, "objects": {"": [[8, 0, 0, "-", "pygenn"]], "pygenn": [[8, 1, 1, "", "CurrentSource"], [8, 1, 1, "", "CustomConnectivityUpdate"], [8, 1, 1, "", "CustomUpdate"], [8, 1, 1, "", "CustomUpdateVarAccess"], [8, 1, 1, "", "CustomUpdateWU"], [8, 1, 1, "", "GeNNModel"], [8, 1, 1, "", "NeuronGroup"], [8, 1, 1, "", "ParallelismHint"], [8, 1, 1, "", "PlogSeverity"], [8, 1, 1, "", "SynapseGroup"], [8, 1, 1, "", "SynapseMatrixConnectivity"], [8, 1, 1, "", "SynapseMatrixType"], [8, 1, 1, "", "SynapseMatrixWeight"], [8, 1, 1, "", "VarAccess"], [8, 1, 1, "", "VarAccessDim"], [8, 1, 1, "", "VarAccessMode"], [8, 1, 1, "", "VarAccessModeAttribute"], [8, 1, 1, "", "VarLocation"], [8, 1, 1, "", "VarLocationAttribute"], [8, 5, 1, "", "create_current_source_model"], [8, 5, 1, "", "create_custom_connectivity_update_model"], [8, 5, 1, "", "create_custom_update_model"], [8, 5, 1, "", "create_egp_ref"], [8, 5, 1, "", "create_neuron_model"], [8, 5, 1, "", "create_post_var_ref"], [8, 5, 1, "", "create_postsynaptic_model"], [8, 5, 1, "", "create_pre_var_ref"], [8, 5, 1, "", "create_psm_egp_ref"], [8, 5, 1, "", "create_psm_var_ref"], [8, 5, 1, "", "create_sparse_connect_init_snippet"], [8, 5, 1, "", "create_toeplitz_connect_init_snippet"], [8, 5, 1, "", "create_var_init_snippet"], [8, 5, 1, "", "create_var_ref"], [8, 5, 1, "", "create_weight_update_model"], [8, 5, 1, "", "create_wu_egp_ref"], [8, 5, 1, "", "create_wu_post_var_ref"], [8, 5, 1, "", "create_wu_pre_var_ref"], [8, 5, 1, "", "create_wu_var_ref"], [8, 0, 0, "-", "cuda_backend"], [8, 0, 0, "-", "current_source_models"], [8, 0, 0, "-", "custom_connectivity_update_models"], [8, 0, 0, "-", "custom_update_models"], [8, 0, 0, "-", "genn_groups"], [8, 5, 1, "", "get_var_access_dim"], [8, 5, 1, "", "init_postsynaptic"], [8, 5, 1, "", "init_sparse_connectivity"], [8, 0, 0, "-", "init_sparse_connectivity_snippets"], [8, 5, 1, "", "init_toeplitz_connectivity"], [8, 0, 0, "-", "init_toeplitz_connectivity_snippets"], [8, 5, 1, "", "init_var"], [8, 0, 0, "-", "init_var_snippets"], [8, 5, 1, "", "init_weight_update"], [8, 0, 0, "-", "model_preprocessor"], [8, 0, 0, "-", "neuron_models"], [8, 0, 0, "-", "postsynaptic_models"], [8, 0, 0, "-", "single_threaded_cpu_backend"], [8, 0, 0, "-", "types"], [8, 0, 0, "-", "weight_update_models"]], "pygenn.CurrentSource": [[8, 2, 1, "", "get_var_location"], [8, 3, 1, "", "model"], [8, 3, 1, "", "name"], [8, 3, 1, "", "params"], [8, 2, 1, "", "set_param_dynamic"], [8, 2, 1, "", "set_var_location"], [8, 3, 1, "", "var_initialisers"]], "pygenn.CustomConnectivityUpdate": [[8, 2, 1, "", "get_post_var_location"], [8, 2, 1, "", "get_pre_var_location"], [8, 2, 1, "", "get_var_location"], [8, 3, 1, "", "model"], [8, 3, 1, "", "name"], [8, 3, 1, "", "params"], [8, 3, 1, "", "post_var_initialisers"], [8, 3, 1, "", "post_var_references"], [8, 3, 1, "", "pre_var_initialisers"], [8, 3, 1, "", "pre_var_references"], [8, 2, 1, "", "set_param_dynamic"], [8, 2, 1, "", "set_post_var_location"], [8, 2, 1, "", "set_pre_var_location"], [8, 2, 1, "", "set_var_location"], [8, 3, 1, "", "synapse_group"], [8, 3, 1, "", "update_group_name"], [8, 3, 1, "", "var_initialisers"], [8, 3, 1, "", "var_references"]], "pygenn.CustomUpdate": [[8, 3, 1, "", "num_neurons"], [8, 3, 1, "", "var_references"]], "pygenn.CustomUpdateVarAccess": [[8, 4, 1, "", "READ_ONLY"], [8, 4, 1, "", "READ_ONLY_SHARED"], [8, 4, 1, "", "READ_ONLY_SHARED_NEURON"], [8, 4, 1, "", "READ_WRITE"], [8, 4, 1, "", "REDUCE_BATCH_MAX"], [8, 4, 1, "", "REDUCE_BATCH_SUM"], [8, 4, 1, "", "REDUCE_NEURON_MAX"], [8, 4, 1, "", "REDUCE_NEURON_SUM"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.CustomUpdateWU": [[8, 3, 1, "", "synapse_group"], [8, 3, 1, "", "var_references"]], "pygenn.GeNNModel": [[8, 2, 1, "", "add_current_source"], [8, 2, 1, "", "add_custom_connectivity_update"], [8, 2, 1, "", "add_custom_update"], [8, 2, 1, "", "add_neuron_population"], [8, 2, 1, "", "add_synapse_population"], [8, 3, 1, "", "backend_name"], [8, 2, 1, "", "build"], [8, 2, 1, "", "custom_update"], [8, 3, 1, "", "dT"], [8, 2, 1, "", "get_custom_update_time"], [8, 2, 1, "", "get_custom_update_transpose_time"], [8, 3, 1, "", "init_sparse_time"], [8, 3, 1, "", "init_time"], [8, 2, 1, "", "load"], [8, 3, 1, "", "neuron_update_time"], [8, 3, 1, "", "postsynaptic_update_time"], [8, 3, 1, "", "presynaptic_update_time"], [8, 2, 1, "", "pull_recording_buffers_from_device"], [8, 2, 1, "", "step_time"], [8, 3, 1, "", "synapse_dynamics_time"], [8, 3, 1, "", "t"], [8, 3, 1, "", "timestep"], [8, 2, 1, "", "unload"]], "pygenn.NeuronGroup": [[8, 2, 1, "", "get_var_location"], [8, 3, 1, "", "model"], [8, 3, 1, "", "name"], [8, 3, 1, "", "num_delay_slots"], [8, 3, 1, "", "num_neurons"], [8, 3, 1, "", "params"], [8, 3, 1, "", "prev_spike_time_location"], [8, 3, 1, "", "prev_spike_time_required"], [8, 3, 1, "", "recording_zero_copy_enabled"], [8, 2, 1, "", "set_param_dynamic"], [8, 2, 1, "", "set_var_location"], [8, 3, 1, "", "spike_event_recording_enabled"], [8, 3, 1, "", "spike_recording_enabled"], [8, 3, 1, "", "spike_time_location"], [8, 3, 1, "", "spike_time_required"], [8, 3, 1, "", "var_initialisers"]], "pygenn.ParallelismHint": [[8, 4, 1, "", "POSTSYNAPTIC"], [8, 4, 1, "", "PRESYNAPTIC"], [8, 4, 1, "", "WORD_PACKED_BITMASK"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.PlogSeverity": [[8, 4, 1, "", "DEBUG"], [8, 4, 1, "", "ERROR"], [8, 4, 1, "", "FATAL"], [8, 4, 1, "", "INFO"], [8, 4, 1, "", "NONE"], [8, 4, 1, "", "VERBOSE"], [8, 4, 1, "", "WARNING"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.SynapseGroup": [[8, 3, 1, "", "axonal_delay_steps"], [8, 3, 1, "", "back_prop_delay_steps"], [8, 3, 1, "", "dendritic_delay_location"], [8, 2, 1, "", "get_ps_var_location"], [8, 2, 1, "", "get_wu_post_var_location"], [8, 2, 1, "", "get_wu_pre_var_location"], [8, 2, 1, "", "get_wu_var_location"], [8, 3, 1, "", "kernel_size"], [8, 3, 1, "", "matrix_type"], [8, 3, 1, "", "max_connections"], [8, 3, 1, "", "max_dendritic_delay_timesteps"], [8, 3, 1, "", "max_source_connections"], [8, 3, 1, "", "name"], [8, 3, 1, "", "narrow_sparse_ind_enabled"], [8, 3, 1, "", "num_threads_per_spike"], [8, 3, 1, "", "output_location"], [8, 3, 1, "", "parallelism_hint"], [8, 3, 1, "", "post_target_var"], [8, 3, 1, "", "pre_target_var"], [8, 3, 1, "", "ps_initialiser"], [8, 2, 1, "", "set_ps_param_dynamic"], [8, 2, 1, "", "set_ps_var_location"], [8, 2, 1, "", "set_wu_param_dynamic"], [8, 2, 1, "", "set_wu_post_var_location"], [8, 2, 1, "", "set_wu_pre_var_location"], [8, 2, 1, "", "set_wu_var_location"], [8, 3, 1, "", "sparse_connectivity_initialiser"], [8, 3, 1, "", "sparse_connectivity_location"], [8, 3, 1, "", "toeplitz_connectivity_initialiser"], [8, 3, 1, "", "wu_initialiser"]], "pygenn.SynapseMatrixConnectivity": [[8, 4, 1, "", "BITMASK"], [8, 4, 1, "", "DENSE"], [8, 4, 1, "", "PROCEDURAL"], [8, 4, 1, "", "SPARSE"], [8, 4, 1, "", "TOEPLITZ"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.SynapseMatrixType": [[8, 4, 1, "", "BITMASK"], [8, 4, 1, "", "DENSE"], [8, 4, 1, "", "DENSE_PROCEDURALG"], [8, 4, 1, "", "PROCEDURAL"], [8, 4, 1, "", "PROCEDURAL_KERNELG"], [8, 4, 1, "", "SPARSE"], [8, 4, 1, "", "TOEPLITZ"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.SynapseMatrixWeight": [[8, 4, 1, "", "INDIVIDUAL"], [8, 4, 1, "", "KERNEL"], [8, 4, 1, "", "PROCEDURAL"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.VarAccess": [[8, 4, 1, "", "READ_ONLY"], [8, 4, 1, "", "READ_ONLY_DUPLICATE"], [8, 4, 1, "", "READ_ONLY_SHARED_NEURON"], [8, 4, 1, "", "READ_WRITE"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.VarAccessDim": [[8, 4, 1, "", "BATCH"], [8, 4, 1, "", "ELEMENT"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.VarAccessMode": [[8, 4, 1, "", "READ_ONLY"], [8, 4, 1, "", "READ_WRITE"], [8, 4, 1, "", "REDUCE_MAX"], [8, 4, 1, "", "REDUCE_SUM"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.VarAccessModeAttribute": [[8, 4, 1, "", "MAX"], [8, 4, 1, "", "READ_ONLY"], [8, 4, 1, "", "READ_WRITE"], [8, 4, 1, "", "REDUCE"], [8, 4, 1, "", "SUM"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.VarLocation": [[8, 4, 1, "", "DEVICE"], [8, 4, 1, "", "HOST_DEVICE"], [8, 4, 1, "", "HOST_DEVICE_ZERO_COPY"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.VarLocationAttribute": [[8, 4, 1, "", "DEVICE"], [8, 4, 1, "", "HOST"], [8, 4, 1, "", "ZERO_COPY"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.cuda_backend": [[8, 1, 1, "", "BlockSizeSelect"], [8, 1, 1, "", "DeviceSelect"], [8, 1, 1, "", "Preferences"]], "pygenn.cuda_backend.BlockSizeSelect": [[8, 4, 1, "", "MANUAL"], [8, 4, 1, "", "OCCUPANCY"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.cuda_backend.DeviceSelect": [[8, 4, 1, "", "MANUAL"], [8, 4, 1, "", "MOST_MEMORY"], [8, 4, 1, "", "OPTIMAL"], [8, 3, 1, "", "name"], [8, 3, 1, "", "value"]], "pygenn.cuda_backend.Preferences": [[8, 3, 1, "", "block_size_select_method"], [8, 3, 1, "", "constant_cache_overhead"], [8, 3, 1, "", "device_select_method"], [8, 3, 1, "", "enable_nccl_reductions"], [8, 3, 1, "", "generate_line_info"], [8, 3, 1, "", "manual_block_sizes"], [8, 3, 1, "", "manual_device_id"], [8, 3, 1, "", "show_ptx_info"]], "pygenn.current_source_models": [[8, 5, 1, "", "DC"], [8, 5, 1, "", "GaussianNoise"], [8, 5, 1, "", "PoissonExp"]], "pygenn.custom_update_models": [[8, 5, 1, "", "Transpose"]], "pygenn.genn_groups": [[8, 1, 1, "", "CurrentSourceMixin"], [8, 1, 1, "", "CustomConnectivityUpdateMixin"], [8, 1, 1, "", "CustomUpdateMixin"], [8, 1, 1, "", "CustomUpdateWUMixin"], [8, 1, 1, "", "GroupMixin"], [8, 1, 1, "", "NeuronGroupMixin"], [8, 1, 1, "", "SynapseGroupMixin"]], "pygenn.genn_groups.CustomConnectivityUpdateMixin": [[8, 2, 1, "", "get_var_values"]], "pygenn.genn_groups.CustomUpdateWUMixin": [[8, 2, 1, "", "get_var_values"]], "pygenn.genn_groups.GroupMixin": [[8, 2, 1, "", "pull_extra_global_param_from_device"], [8, 2, 1, "", "pull_var_from_device"], [8, 2, 1, "", "push_extra_global_param_to_device"], [8, 2, 1, "", "push_var_to_device"], [8, 2, 1, "", "set_dynamic_param_value"]], "pygenn.genn_groups.NeuronGroupMixin": [[8, 3, 1, "", "spike_recording_data"]], "pygenn.genn_groups.SynapseGroupMixin": [[8, 2, 1, "", "get_sparse_post_inds"], [8, 2, 1, "", "get_sparse_pre_inds"], [8, 2, 1, "", "get_var_values"], [8, 3, 1, "", "post_spike_event_recording_data"], [8, 3, 1, "", "pre_spike_event_recording_data"], [8, 2, 1, "", "pull_connectivity_from_device"], [8, 2, 1, "", "pull_in_syn_from_device"], [8, 2, 1, "", "pull_psm_extra_global_param_from_device"], [8, 2, 1, "", "push_connectivity_to_device"], [8, 2, 1, "", "push_in_syn_to_device"], [8, 2, 1, "", "push_psm_extra_global_param_to_device"], [8, 2, 1, "", "set_sparse_connections"], [8, 3, 1, "", "synapse_group"], [8, 3, 1, "", "weight_update_var_size"]], "pygenn.init_sparse_connectivity_snippets": [[8, 5, 1, "", "Conv2D"], [8, 5, 1, "", "FixedNumberPostWithReplacement"], [8, 5, 1, "", "FixedNumberPreWithReplacement"], [8, 5, 1, "", "FixedNumberTotalWithReplacement"], [8, 5, 1, "", "FixedProbability"], [8, 5, 1, "", "FixedProbabilityNoAutapse"], [8, 5, 1, "", "OneToOne"], [8, 5, 1, "", "Uninitialised"]], "pygenn.init_toeplitz_connectivity_snippets": [[8, 5, 1, "", "AvgPoolConv2D"], [8, 5, 1, "", "Conv2D"], [8, 5, 1, "", "Uninitialised"]], "pygenn.init_var_snippets": [[8, 5, 1, "", "Binomial"], [8, 5, 1, "", "Constant"], [8, 5, 1, "", "Exponential"], [8, 5, 1, "", "Gamma"], [8, 5, 1, "", "Kernel"], [8, 5, 1, "", "Normal"], [8, 5, 1, "", "NormalClipped"], [8, 5, 1, "", "NormalClippedDelay"], [8, 5, 1, "", "Uniform"], [8, 5, 1, "", "Uninitialised"]], "pygenn.model_preprocessor": [[8, 1, 1, "", "Array"], [8, 1, 1, "", "ArrayBase"], [8, 1, 1, "", "ExtraGlobalParameter"], [8, 1, 1, "", "SynapseVariable"], [8, 1, 1, "", "Variable"], [8, 1, 1, "", "VariableBase"]], "pygenn.model_preprocessor.Array": [[8, 3, 1, "", "view"]], "pygenn.model_preprocessor.ArrayBase": [[8, 2, 1, "", "pull_from_device"], [8, 2, 1, "", "push_to_device"], [8, 2, 1, "", "set_array"]], "pygenn.model_preprocessor.ExtraGlobalParameter": [[8, 2, 1, "", "set_init_values"], [8, 2, 1, "", "set_values"], [8, 3, 1, "", "values"], [8, 3, 1, "", "view"]], "pygenn.model_preprocessor.SynapseVariable": [[8, 3, 1, "", "current_values"], [8, 3, 1, "", "current_view"], [8, 3, 1, "", "values"], [8, 3, 1, "", "view"]], "pygenn.model_preprocessor.Variable": [[8, 3, 1, "", "current_values"], [8, 3, 1, "", "current_view"], [8, 3, 1, "", "values"], [8, 3, 1, "", "view"]], "pygenn.model_preprocessor.VariableBase": [[8, 2, 1, "", "set_array"], [8, 2, 1, "", "set_init_values"], [8, 2, 1, "", "set_values"]], "pygenn.neuron_models": [[8, 5, 1, "", "Izhikevich"], [8, 5, 1, "", "IzhikevichVariable"], [8, 5, 1, "", "LIF"], [8, 5, 1, "", "Poisson"], [8, 5, 1, "", "PoissonNew"], [8, 5, 1, "", "RulkovMap"], [8, 5, 1, "", "SpikeSource"], [8, 5, 1, "", "SpikeSourceArray"], [8, 5, 1, "", "TraubMiles"], [8, 5, 1, "", "TraubMilesAlt"], [8, 5, 1, "", "TraubMilesFast"], [8, 5, 1, "", "TraubMilesNStep"]], "pygenn.postsynaptic_models": [[8, 5, 1, "", "DeltaCurr"], [8, 5, 1, "", "ExpCond"], [8, 5, 1, "", "ExpCurr"]], "pygenn.single_threaded_cpu_backend": [[8, 1, 1, "", "Preferences"]], "pygenn.weight_update_models": [[8, 5, 1, "", "PiecewiseSTDP"], [8, 5, 1, "", "StaticGraded"], [8, 5, 1, "", "StaticPulse"], [8, 5, 1, "", "StaticPulseConstantWeight"], [8, 5, 1, "", "StaticPulseDendriticDelay"]]}, "objtypes": {"0": "py:module", "1": "py:class", "2": "py:method", "3": "py:property", "4": "py:attribute", "5": "py:function"}, "objnames": {"0": ["py", "module", "Python module"], "1": ["py", "class", "Python class"], "2": ["py", "method", "Python method"], "3": ["py", "property", "Python property"], "4": ["py", "attribute", "Python attribute"], "5": ["py", "function", "Python function"]}, "titleterms": {"bibliographi": 0, "build": [1, 4], "network": [1, 6], "The": 1, "model": [1, 2, 11, 12], "popul": 1, "paramet": [1, 6], "extra": [1, 6], "global": [1, 6], "refer": 1, "variabl": [1, 2, 6], "locat": 1, "neuron": [1, 2], "synaps": 1, "current": [1, 2], "sourc": [1, 2], "custom": [1, 2], "updat": [1, 2], "connect": [1, 2], "genncod": [2, 9], "random": 2, "number": 2, "gener": 2, "initialis": 2, "snippet": 2, "spars": 2, "toeplitz": 2, "weight": 2, "postsynapt": 2, "pygenn": [3, 7, 8, 12, 14], "document": 3, "indic": 3, "tabl": 3, "instal": 4, "pre": 4, "pip": 4, "setup": 4, "py": 4, "comput": [5, 13], "time": [5, 13], "simul": 6, "spike": 6, "record": 6, "push": 6, "pull": 6, "valu": 6, "view": 6, "dynam": 6, "packag": 8, "submodul": 8, "cuda_backend": 8, "modul": 8, "current_source_model": 8, "custom_connectivity_update_model": 8, "custom_update_model": 8, "genn": [8, 9], "genn_group": 8, "init_sparse_connectivity_snippet": 8, "init_toeplitz_connectivity_snippet": 8, "init_var_snippet": 8, "model_preprocessor": 8, "neuron_model": 8, "postsynaptic_model": 8, "runtim": 8, "single_threaded_cpu_backend": 8, "type": 8, "weight_update_model": 8, "upgrad": 9, "from": 9, "4": 9, "syntax": 9, "chang": 9, "user": 10, "project": 10, "mnist": 11, "classif": 11, "us": 11, "an": 11, "insect": 11, "inspir": 11, "mushroom": 11, "bodi": 11, "name": [11, 12, 14], "argument": [11, 12, 14], "implement": [12, 14], "local": 12, "cortic": 12, "microcircuit": 12, "superspik": 14}, "envversion": {"sphinx.domains.c": 2, "sphinx.domains.changeset": 1, "sphinx.domains.citation": 1, "sphinx.domains.cpp": 6, "sphinx.domains.index": 1, "sphinx.domains.javascript": 2, "sphinx.domains.math": 2, "sphinx.domains.python": 3, "sphinx.domains.rst": 2, "sphinx.domains.std": 2, "sphinx": 56}})
\ No newline at end of file
diff --git a/documentation/5/source/modules.html b/documentation/5/source/modules.html
index b13f7a9d4..0fe06b190 100644
--- a/documentation/5/source/modules.html
+++ b/documentation/5/source/modules.html
@@ -52,6 +52,7 @@ <h1>pygenn<a class="headerlink" href="#pygenn" title="Permalink to this heading"
 <li class="toctree-l2"><a class="reference internal" href="pygenn.html#module-pygenn.current_source_models">pygenn.current_source_models module</a></li>
 <li class="toctree-l2"><a class="reference internal" href="pygenn.html#module-pygenn.custom_connectivity_update_models">pygenn.custom_connectivity_update_models module</a></li>
 <li class="toctree-l2"><a class="reference internal" href="pygenn.html#module-pygenn.custom_update_models">pygenn.custom_update_models module</a></li>
+<li class="toctree-l2"><a class="reference internal" href="pygenn.html#pygenn-genn-module">pygenn.genn module</a></li>
 <li class="toctree-l2"><a class="reference internal" href="pygenn.html#module-pygenn.genn_groups">pygenn.genn_groups module</a></li>
 <li class="toctree-l2"><a class="reference internal" href="pygenn.html#module-pygenn.init_sparse_connectivity_snippets">pygenn.init_sparse_connectivity_snippets module</a></li>
 <li class="toctree-l2"><a class="reference internal" href="pygenn.html#module-pygenn.init_toeplitz_connectivity_snippets">pygenn.init_toeplitz_connectivity_snippets module</a></li>
diff --git a/documentation/5/source/pygenn.html b/documentation/5/source/pygenn.html
index fd4bed223..7c1f15f65 100644
--- a/documentation/5/source/pygenn.html
+++ b/documentation/5/source/pygenn.html
@@ -314,15 +314,16 @@ <h3>Navigation</h3>
 <dl class="field-list simple">
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
-<li><p><strong>precision</strong> (<em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em>) – </p></li>
-<li><p><strong>model_name</strong> (<em>str</em>) – </p></li>
-<li><p><strong>backend</strong> (<em>Optional</em><em>[</em><em>str</em><em>]</em>) – </p></li>
-<li><p><strong>time_precision</strong> (<em>Optional</em><em>[</em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>]</em>) – </p></li>
-<li><p><strong>genn_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – </p></li>
-<li><p><strong>code_gen_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – </p></li>
-<li><p><strong>transpiler_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – </p></li>
-<li><p><strong>runtime_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – </p></li>
-<li><p><strong>backend_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – </p></li>
+<li><p><strong>precision</strong> (<em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em>) – Data type to use for <code class="docutils literal notranslate"><span class="pre">scalar</span></code> variables</p></li>
+<li><p><strong>model_name</strong> (<em>str</em>) – Name of the model</p></li>
+<li><p><strong>backend</strong> (<em>Optional</em><em>[</em><em>str</em><em>]</em>) – Name of backend module to use. Defaults to one to pick ‘best’ backend for your system</p></li>
+<li><p><strong>time_precision</strong> (<em>Optional</em><em>[</em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>]</em>) – data type to use for representing time</p></li>
+<li><p><strong>genn_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – Log level for GeNN</p></li>
+<li><p><strong>code_gen_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – Log level for GeNN code-generator</p></li>
+<li><p><strong>transpiler_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – Log level for GeNN transpiler</p></li>
+<li><p><strong>runtime_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – Log level for GeNN runtime</p></li>
+<li><p><strong>backend_log_level</strong> (<a class="reference internal" href="#pygenn.PlogSeverity" title="pygenn._genn.PlogSeverity"><em>PlogSeverity</em></a>) – Log level for backend</p></li>
+<li><p><strong>preference_kwargs</strong> – Additional keyword arguments to set in backend preferences structure</p></li>
 </ul>
 </dd>
 </dl>
@@ -334,7 +335,7 @@ <h3>Navigation</h3>
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>cs_name</strong> (<em>str</em>) – unique name</p></li>
-<li><p><strong>current_source_model</strong> (<em>Union</em><em>[</em><em>CurrentSourceModelBase</em><em>, </em><em>str</em><em>]</em>) – current source model either as a string referencing a built in model
+<li><p><strong>current_source_model</strong> (<em>Union</em><em>[</em><em>CurrentSourceModelBase</em><em>, </em><em>str</em><em>]</em>) – current source model either as a string referencing a built-in model
 (see <a class="reference internal" href="#module-pygenn.current_source_models" title="pygenn.current_source_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">current_source_models</span></code></a>) or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">CurrentSourceModelBase</span></code>
 (for example returned by <a class="reference internal" href="#pygenn.create_current_source_model" title="pygenn.create_current_source_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_current_source_model()</span></code></a>)</p></li>
 <li><p><strong>pop</strong> (<a class="reference internal" href="#pygenn.NeuronGroup" title="pygenn._genn.NeuronGroup"><em>NeuronGroup</em></a>) – neuron population to inject current into</p></li>
@@ -361,7 +362,7 @@ <h3>Navigation</h3>
 
 <dl class="py method">
 <dt class="sig sig-object py" id="pygenn.GeNNModel.add_custom_connectivity_update">
-<span class="sig-name descname"><span class="pre">add_custom_connectivity_update</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">cu_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">group_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">syn_group</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_conn_update_model</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#pygenn.GeNNModel.add_custom_connectivity_update" title="Permalink to this definition">¶</a></dt>
+<span class="sig-name descname"><span class="pre">add_custom_connectivity_update</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">cu_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">group_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">syn_group</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">custom_conn_update_model</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">egp_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#pygenn.GeNNModel.add_custom_connectivity_update" title="Permalink to this definition">¶</a></dt>
 <dd><p>Add a custom connectivity update to the GeNN model</p>
 <dl class="field-list simple">
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
@@ -370,7 +371,7 @@ <h3>Navigation</h3>
 <li><p><strong>group_name</strong> (<em>str</em>) – name of the ‘custom update group’ to include this update in.
 All custom updates in the same group are executed simultaneously.</p></li>
 <li><p><strong>syn_group</strong> (<a class="reference internal" href="#pygenn.SynapseGroup" title="pygenn._genn.SynapseGroup"><em>SynapseGroup</em></a>) – Synapse group to attach custom connectivity update to</p></li>
-<li><p><strong>custom_conn_update_model</strong> (<em>Union</em><em>[</em><em>CustomConnectivityUpdateModelBase</em><em>, </em><em>str</em><em>]</em>) – custom connectivity update model either as a string referencing a built in model
+<li><p><strong>custom_conn_update_model</strong> (<em>Union</em><em>[</em><em>CustomConnectivityUpdateModelBase</em><em>, </em><em>str</em><em>]</em>) – custom connectivity update model either as a string referencing a built-in model
 (see <a class="reference internal" href="#module-pygenn.custom_connectivity_update_models" title="pygenn.custom_connectivity_update_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">custom_connectivity_update_models</span></code></a>) or an instance of
 <code class="xref py py-class docutils literal notranslate"><span class="pre">CustomConnectivityUpdateModelBaseUpdateModelBase</span></code>
 (for example returned by <a class="reference internal" href="#pygenn.create_custom_connectivity_update_model" title="pygenn.create_custom_connectivity_update_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_custom_connectivity_update_model()</span></code></a>)</p></li>
@@ -390,6 +391,9 @@ <h3>Navigation</h3>
 <li><p><strong>post_var_refs</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>VarReference</em><em>]</em>) – references to postsynaptic variables,
 typically created using <a class="reference internal" href="#pygenn.create_var_ref" title="pygenn.create_var_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_var_ref()</span></code></a>
 (see <a href="#id30"><span class="problematic" id="id31">`Variables references`_</span></a>)</p></li>
+<li><p><strong>egp_refs</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>EGPReference</em><em>]</em>) – references to extra global parameters in other populations
+to access from this update, typically created using
+<a class="reference internal" href="#pygenn.create_egp_ref" title="pygenn.create_egp_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_egp_ref()</span></code></a> (see <a href="#id32"><span class="problematic" id="id33">`Extra global parameter references`_</span></a>).</p></li>
 </ul>
 </dd>
 </dl>
@@ -405,20 +409,20 @@ <h3>Navigation</h3>
 <li><p><strong>cu_name</strong> (<em>str</em>) – unique name</p></li>
 <li><p><strong>group_name</strong> (<em>str</em>) – name of the ‘custom update group’ to include this update in.
 All custom updates in the same group are executed simultaneously.</p></li>
-<li><p><strong>custom_update_model</strong> (<em>Union</em><em>[</em><em>CustomUpdateModelBase</em><em>, </em><em>str</em><em>]</em>) – custom update model either as a string referencing a built in model
+<li><p><strong>custom_update_model</strong> (<em>Union</em><em>[</em><em>CustomUpdateModelBase</em><em>, </em><em>str</em><em>]</em>) – custom update model either as a string referencing a built-in model
 (see <a class="reference internal" href="#module-pygenn.custom_update_models" title="pygenn.custom_update_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">custom_update_models</span></code></a>) or an instance of
 <code class="xref py py-class docutils literal notranslate"><span class="pre">CustomUpdateModelBase</span></code>
 (for example returned by <a class="reference internal" href="#pygenn.create_custom_update_model" title="pygenn.create_custom_update_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_custom_update_model()</span></code></a>)</p></li>
-<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the custom update model (see <a href="#id32"><span class="problematic" id="id33">`Parameters`_</span></a>)</p></li>
+<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the custom update model (see <a href="#id34"><span class="problematic" id="id35">`Parameters`_</span></a>)</p></li>
 <li><p><strong>vars</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>VarInit</em><em>, </em><em>int</em><em>, </em><em>float</em><em>, </em><em>ndarray</em><em>, </em><em>Sequence</em><em>]</em><em>]</em>) – initial variable values or initialisers
-for the custom update model (see <a href="#id34"><span class="problematic" id="id35">`Variables`_</span></a>)</p></li>
+for the custom update model (see <a href="#id36"><span class="problematic" id="id37">`Variables`_</span></a>)</p></li>
 <li><p><strong>var_refs</strong> (<em>Union</em><em>[</em><em>Dict</em><em>[</em><em>str</em><em>, </em><em>VarReference</em><em>]</em><em>, </em><em>Dict</em><em>[</em><em>str</em><em>, </em><em>WUVarReference</em><em>]</em><em>]</em>) – references to variables in other populations to
 access from this update, typically created using either
 <a class="reference internal" href="#pygenn.create_var_ref" title="pygenn.create_var_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_var_ref()</span></code></a> or <a class="reference internal" href="#pygenn.create_wu_var_ref" title="pygenn.create_wu_var_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_wu_var_ref()</span></code></a>
-(see <a href="#id36"><span class="problematic" id="id37">`Variables references`_</span></a>).</p></li>
+(see <a href="#id38"><span class="problematic" id="id39">`Variables references`_</span></a>).</p></li>
 <li><p><strong>egp_refs</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>EGPReference</em><em>]</em>) – references to extra global parameters in other populations
 to access from this update, typically created using
-<a class="reference internal" href="#pygenn.create_egp_ref" title="pygenn.create_egp_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_egp_ref()</span></code></a> (see <a href="#id38"><span class="problematic" id="id39">`Extra global parameter references`_</span></a>).</p></li>
+<a class="reference internal" href="#pygenn.create_egp_ref" title="pygenn.create_egp_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_egp_ref()</span></code></a> (see <a href="#id40"><span class="problematic" id="id41">`Extra global parameter references`_</span></a>).</p></li>
 </ul>
 </dd>
 </dl>
@@ -433,19 +437,19 @@ <h3>Navigation</h3>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>pop_name</strong> (<em>str</em>) – unique name</p></li>
 <li><p><strong>num_neurons</strong> (<em>int</em>) – number of neurons</p></li>
-<li><p><strong>neuron</strong> (<em>Union</em><em>[</em><em>NeuronModelBase</em><em>, </em><em>str</em><em>]</em>) – neuron model either as a string referencing a built in model
+<li><p><strong>neuron</strong> (<em>Union</em><em>[</em><em>NeuronModelBase</em><em>, </em><em>str</em><em>]</em>) – neuron model either as a string referencing a built-in model
 (see <a class="reference internal" href="#module-pygenn.neuron_models" title="pygenn.neuron_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">neuron_models</span></code></a>) or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">NeuronModelBase</span></code>
 (for example returned by <a class="reference internal" href="#pygenn.create_neuron_model" title="pygenn.create_neuron_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_neuron_model()</span></code></a>)</p></li>
-<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the neuron model (see <a href="#id40"><span class="problematic" id="id41">`Parameters`_</span></a>)</p></li>
+<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the neuron model (see <a href="#id42"><span class="problematic" id="id43">`Parameters`_</span></a>)</p></li>
 <li><p><strong>vars</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>VarInit</em><em>, </em><em>int</em><em>, </em><em>float</em><em>, </em><em>ndarray</em><em>, </em><em>Sequence</em><em>]</em><em>]</em>) – initial variable values or initialisers
-for the neuron model (see <a href="#id42"><span class="problematic" id="id43">`Variables`_</span></a>)</p></li>
+for the neuron model (see <a href="#id44"><span class="problematic" id="id45">`Variables`_</span></a>)</p></li>
 </ul>
 </dd>
 <dt class="field-even">Return type<span class="colon">:</span></dt>
 <dd class="field-even"><p><a class="reference internal" href="#pygenn.NeuronGroup" title="pygenn._genn.NeuronGroup"><em>NeuronGroup</em></a></p>
 </dd>
 </dl>
-<p>For example, a population of 10 neurons using the built in Izhikevich model and
+<p>For example, a population of 10 neurons using the built-in Izhikevich model and
 the standard set of ‘tonic spiking’ parameters could be added to a model as follows:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">pop</span> <span class="o">=</span> <span class="n">model</span><span class="o">.</span><span class="n">add_neuron_population</span><span class="p">(</span><span class="s2">&quot;pop&quot;</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="s2">&quot;Izhikevich&quot;</span><span class="p">,</span>
                                   <span class="p">{</span><span class="s2">&quot;a&quot;</span><span class="p">:</span> <span class="mf">0.02</span><span class="p">,</span> <span class="s2">&quot;b&quot;</span><span class="p">:</span> <span class="mf">0.2</span><span class="p">,</span> <span class="s2">&quot;c&quot;</span><span class="p">:</span> <span class="o">-</span><span class="mf">65.0</span><span class="p">,</span> <span class="s2">&quot;d&quot;</span><span class="p">:</span> <span class="mf">6.0</span><span class="p">},</span>
@@ -1406,7 +1410,7 @@ <h3>Navigation</h3>
 Within the <code class="docutils literal notranslate"><span class="pre">injection_code</span></code> code string, the variables, parameters,
 derived parameters, neuron variable references and extra global
 parameters defined in this model can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">id</span></code> which represents a neurons index within a population (starting from zero)</p></li>
@@ -1446,7 +1450,7 @@ <h3>Navigation</h3>
 
 <dl class="py function">
 <dt class="sig sig-object py" id="pygenn.create_custom_connectivity_update_model">
-<span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">create_custom_connectivity_update_model</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">class_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">derived_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">row_update_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">host_update_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">extra_global_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#pygenn.create_custom_connectivity_update_model" title="Permalink to this definition">¶</a></dt>
+<span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">create_custom_connectivity_update_model</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">class_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">derived_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pre_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">post_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">row_update_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">host_update_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">extra_global_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">extra_global_param_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#pygenn.create_custom_connectivity_update_model" title="Permalink to this definition">¶</a></dt>
 <dd><p>Creates a new custom connectivity update model.</p>
 <p>Within host update code, you have full access to parameters, derived parameters,
 extra global parameters and pre and postsynaptic variables. By design you do
@@ -1482,6 +1486,8 @@ <h3>Navigation</h3>
 on CPU when custom connectivity update is launched</p></li>
 <li><p><strong>extra_global_params</strong> – names and types of model
 extra global parameters</p></li>
+<li><p><strong>extra_global_param_refs</strong> – names and types of extra global
+parameter references</p></li>
 <li><p><strong>post_vars</strong> (<em>Optional</em><em>[</em><em>Sequence</em><em>[</em><em>Union</em><em>[</em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>]</em><em>, </em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>, </em><a class="reference internal" href="#pygenn.VarAccess" title="pygenn._genn.VarAccess"><em>VarAccess</em></a><em>]</em><em>]</em><em>]</em><em>]</em>) – </p></li>
 <li><p><strong>pre_var_refs</strong> (<em>Optional</em><em>[</em><em>Sequence</em><em>[</em><em>Union</em><em>[</em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>]</em><em>, </em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>, </em><a class="reference internal" href="#pygenn.VarAccessMode" title="pygenn._genn.VarAccessMode"><em>VarAccessMode</em></a><em>]</em><em>]</em><em>]</em><em>]</em>) – </p></li>
 <li><p><strong>post_var_refs</strong> (<em>Optional</em><em>[</em><em>Sequence</em><em>[</em><em>Union</em><em>[</em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>]</em><em>, </em><em>Tuple</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>str</em><em>, </em><em>ResolvedType</em><em>]</em><em>, </em><a class="reference internal" href="#pygenn.VarAccessMode" title="pygenn._genn.VarAccessMode"><em>VarAccessMode</em></a><em>]</em><em>]</em><em>]</em><em>]</em>) – </p></li>
@@ -1552,7 +1558,7 @@ <h3>Navigation</h3>
 Within the <code class="docutils literal notranslate"><span class="pre">update_code</span></code> code string, the variables, parameters,
 derived parameters, variable references, extra global parameters
 and extra global parameter references defined in this model can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 </ul>
@@ -1646,6 +1652,7 @@ <h3>Navigation</h3>
 <li><p>create_egp_ref(arg0: GeNN::CurrentSource, arg1: str) -&gt; GeNN::Models::EGPReference</p></li>
 <li><p>create_egp_ref(arg0: GeNN::CustomUpdate, arg1: str) -&gt; GeNN::Models::EGPReference</p></li>
 <li><p>create_egp_ref(arg0: GeNN::CustomUpdateWU, arg1: str) -&gt; GeNN::Models::EGPReference</p></li>
+<li><p>create_egp_ref(arg0: GeNN::CustomConnectivityUpdate, arg1: str) -&gt; GeNN::Models::EGPReference</p></li>
 </ol>
 </dd></dl>
 
@@ -1656,7 +1663,7 @@ <h3>Navigation</h3>
 Within all of the code strings, the variables, parameters,
 derived parameters, additional input variables and extra global
 parameters defined in this model can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">Isyn</span></code> which represents the total incoming synaptic input.</p></li>
@@ -1720,6 +1727,11 @@ <h3>Navigation</h3>
 </div>
 </dd></dl>
 
+<dl class="py function">
+<dt class="sig sig-object py" id="pygenn.create_post_var_ref">
+<span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">create_post_var_ref</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">arg0:</span> <span class="pre">GeNN::CustomConnectivityUpdate</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">arg1:</span> <span class="pre">str</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">GeNN::Models::VarReference</span></span></span><a class="headerlink" href="#pygenn.create_post_var_ref" title="Permalink to this definition">¶</a></dt>
+<dd></dd></dl>
+
 <dl class="py function">
 <dt class="sig sig-object py" id="pygenn.create_postsynaptic_model">
 <span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">create_postsynaptic_model</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">class_name</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">vars</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">neuron_var_refs</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">derived_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">sim_code</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">extra_global_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="headerlink" href="#pygenn.create_postsynaptic_model" title="Permalink to this definition">¶</a></dt>
@@ -1727,7 +1739,7 @@ <h3>Navigation</h3>
 Within all of the code strings, the variables, parameters,
 derived parameters and extra global parameters defined in this model
 can all be referred to be name. Additionally, the code may refer to the
-following built in read-only variables:</p>
+following built-in read-only variables:</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">id</span></code> which represents a neurons index within a population (starting from zero)</p></li>
@@ -1757,6 +1769,11 @@ <h3>Navigation</h3>
 </dl>
 </dd></dl>
 
+<dl class="py function">
+<dt class="sig sig-object py" id="pygenn.create_pre_var_ref">
+<span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">create_pre_var_ref</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">arg0:</span> <span class="pre">GeNN::CustomConnectivityUpdate</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">arg1:</span> <span class="pre">str</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">GeNN::Models::VarReference</span></span></span><a class="headerlink" href="#pygenn.create_pre_var_ref" title="Permalink to this definition">¶</a></dt>
+<dd></dd></dl>
+
 <dl class="py function">
 <dt class="sig sig-object py" id="pygenn.create_psm_egp_ref">
 <span class="sig-prename descclassname"><span class="pre">pygenn.</span></span><span class="sig-name descname"><span class="pre">create_psm_egp_ref</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">arg0:</span> <span class="pre">GeNN::SynapseGroup</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">arg1:</span> <span class="pre">str</span></span></em><span class="sig-paren">)</span> <span class="sig-return"><span class="sig-return-icon">&#x2192;</span> <span class="sig-return-typehint"><span class="pre">GeNN::Models::EGPReference</span></span></span><a class="headerlink" href="#pygenn.create_psm_egp_ref" title="Permalink to this definition">¶</a></dt>
@@ -1773,7 +1790,7 @@ <h3>Navigation</h3>
 <dd><p>Creates a new sparse connectivity initialisation snippet
 Within the code strings, the parameters, derived parameters and
 extra global parameters defined in this snippet can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">num_pre</span></code> which represents the number of presynaptic neurons</p></li>
@@ -1781,14 +1798,16 @@ <h3>Navigation</h3>
 <li><p><code class="docutils literal notranslate"><span class="pre">thread</span></code> when some procedural connectivity is used with multiple
 threads per presynaptic neuron, represents the index of thread</p></li>
 </ul>
-<p>and, in <code class="docutils literal notranslate"><span class="pre">row_build_code</span></code>:
-- <code class="docutils literal notranslate"><span class="pre">id_pre</span></code> represents the index of the presynaptic neuron (starting from zero)
-- <code class="docutils literal notranslate"><span class="pre">id_post_begin</span></code> when some procedural connectivity is used with multiple</p>
-<blockquote>
-<div><p>threads per presynaptic neuron, represents the index of the first postsynaptic neuron to connect.</p>
-</div></blockquote>
-<p>and, in <code class="docutils literal notranslate"><span class="pre">col_build_code</span></code>:
-- <code class="docutils literal notranslate"><span class="pre">id_post</span></code> which represents the index of the postsynaptic neuron (starting from zero)</p>
+<p>and, in <code class="docutils literal notranslate"><span class="pre">row_build_code</span></code>:</p>
+<ul class="simple">
+<li><p><code class="docutils literal notranslate"><span class="pre">id_pre</span></code> represents the index of the presynaptic neuron (starting from zero)</p></li>
+<li><p><code class="docutils literal notranslate"><span class="pre">id_post_begin</span></code> when some procedural connectivity is used with multiple
+threads per presynaptic neuron, represents the index of the first postsynaptic neuron to connect.</p></li>
+</ul>
+<p>and, in <code class="docutils literal notranslate"><span class="pre">col_build_code</span></code>:</p>
+<ul class="simple">
+<li><p><code class="docutils literal notranslate"><span class="pre">id_post</span></code> which represents the index of the postsynaptic neuron (starting from zero)</p></li>
+</ul>
 <p>Finally, the function <code class="docutils literal notranslate"><span class="pre">addSynapse(x)</span></code> can be used to add a new synapse to the connectivity
 where, in <code class="docutils literal notranslate"><span class="pre">row_build_code</span></code>, <code class="docutils literal notranslate"><span class="pre">x</span></code> is the index of the postsynaptic neuron to connect <cite>id_pre`</cite> to
 and, in <code class="docutils literal notranslate"><span class="pre">col_build_code</span></code>, <code class="docutils literal notranslate"><span class="pre">x</span></code> is the index of the presynaptic neuron to connect to <code class="docutils literal notranslate"><span class="pre">id_post</span></code></p>
@@ -1846,7 +1865,7 @@ <h3>Navigation</h3>
 snippet of code specified using the <cite>diagonal_build_code`</cite>.
 Within the code strings, the parameters, derived parameters and
 extra global parameters defined in this snippet can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">num_pre</span></code> which represents the number of presynaptic neurons</p></li>
@@ -1910,7 +1929,7 @@ <h3>Navigation</h3>
 <dd><p>Creates a new variable initialisation snippet
 Within the <code class="docutils literal notranslate"><span class="pre">var_init_code</span></code>, the parameters, derived parameters and
 extra global parameters defined in this snippet can all be referred to be name.
-Additionally, the code may refer to the following built in read-only variables</p>
+Additionally, the code may refer to the following built-in read-only variables</p>
 <ul class="simple">
 <li><p><code class="docutils literal notranslate"><span class="pre">dt</span></code> which represents the simulation time step (as specified via  <code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.dt()</span></code>)</p></li>
 </ul>
@@ -2187,20 +2206,20 @@ <h3>Navigation</h3>
 <dl class="field-list simple">
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
-<li><p><strong>snippet</strong> (<em>Union</em><em>[</em><em>PostsynapticModelBase</em><em>, </em><em>str</em><em>]</em>) – postsynaptic model either as a string referencing a built in model
+<li><p><strong>snippet</strong> (<em>Union</em><em>[</em><em>PostsynapticModelBase</em><em>, </em><em>str</em><em>]</em>) – postsynaptic model either as a string referencing a built-in model
 (see <code class="xref py py-mod docutils literal notranslate"><span class="pre">postsynaptic_models_models</span></code>) or an instance of
 <code class="xref py py-class docutils literal notranslate"><span class="pre">PostsynapticModelBase</span></code> (for example returned
 by <a class="reference internal" href="#pygenn.create_postsynaptic_model" title="pygenn.create_postsynaptic_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_postsynaptic_model()</span></code></a>)</p></li>
-<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the postsynaptic model (see <a href="#id44"><span class="problematic" id="id45">`Parameters`_</span></a>)</p></li>
+<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the postsynaptic model (see <a href="#id46"><span class="problematic" id="id47">`Parameters`_</span></a>)</p></li>
 <li><p><strong>vars</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>VarInit</em><em>, </em><em>int</em><em>, </em><em>float</em><em>, </em><em>ndarray</em><em>, </em><em>Sequence</em><em>]</em><em>]</em>) – initial synaptic variable values or initialisers
-for the postsynaptic model (see <a href="#id46"><span class="problematic" id="id47">`Variables`_</span></a>)</p></li>
+for the postsynaptic model (see <a href="#id48"><span class="problematic" id="id49">`Variables`_</span></a>)</p></li>
 <li><p><strong>var_refs</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>VarInit</em><em>, </em><em>int</em><em>, </em><em>float</em><em>, </em><em>ndarray</em><em>, </em><em>Sequence</em><em>]</em><em>]</em>) – references to postsynaptic neuron variables,
 typically created using <a class="reference internal" href="#pygenn.create_var_ref" title="pygenn.create_var_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_var_ref()</span></code></a>
-(see <a href="#id48"><span class="problematic" id="id49">`Variables references`_</span></a>)</p></li>
+(see <a href="#id50"><span class="problematic" id="id51">`Variables references`_</span></a>)</p></li>
 </ul>
 </dd>
 </dl>
-<p>For example, the built in conductance model with exponential
+<p>For example, the built-in conductance model with exponential
 current shaping could be initialised as follows:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">postsynaptic_init</span> <span class="o">=</span> <span class="n">init_postsynaptic</span><span class="p">(</span><span class="s2">&quot;ExpCond&quot;</span><span class="p">,</span> <span class="p">{</span><span class="s2">&quot;tau&quot;</span><span class="p">:</span> <span class="mf">1.0</span><span class="p">,</span> <span class="s2">&quot;E&quot;</span><span class="p">:</span> <span class="o">-</span><span class="mf">80.0</span><span class="p">},</span>
                                       <span class="n">var_refs</span><span class="o">=</span><span class="p">{</span><span class="s2">&quot;V&quot;</span><span class="p">:</span> <span class="n">create_var_ref</span><span class="p">(</span><span class="n">pop1</span><span class="p">,</span> <span class="s2">&quot;V&quot;</span><span class="p">)})</span>
@@ -2219,14 +2238,14 @@ <h3>Navigation</h3>
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>snippet</strong> (<em>Union</em><em>[</em><em>InitSparseConnectivitySnippetBase</em><em>, </em><em>str</em><em>]</em>) – sparse connectivity init snippet, either as a string referencing
-a built in snipept (see <a class="reference internal" href="#module-pygenn.init_sparse_connectivity_snippets" title="pygenn.init_sparse_connectivity_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_sparse_connectivity_snippets</span></code></a>)
+a built-in snippet (see <a class="reference internal" href="#module-pygenn.init_sparse_connectivity_snippets" title="pygenn.init_sparse_connectivity_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_sparse_connectivity_snippets</span></code></a>)
 or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">InitSparseConnectivitySnippetBase</span></code>
 (for example returned by <a class="reference internal" href="#pygenn.create_sparse_connect_init_snippet" title="pygenn.create_sparse_connect_init_snippet"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_sparse_connect_init_snippet()</span></code></a>)</p></li>
-<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the sparse connectivity init snippet (see <a href="#id50"><span class="problematic" id="id51">`Parameters`_</span></a>)</p></li>
+<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the sparse connectivity init snippet (see <a href="#id52"><span class="problematic" id="id53">`Parameters`_</span></a>)</p></li>
 </ul>
 </dd>
 </dl>
-<p>For example, the built in model could be initialised to generate connectivity
+<p>For example, the built-in model could be initialised to generate connectivity
 where each pair of pre and postsynaptic neurons is connected with a probability of 0.1:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">init</span> <span class="o">=</span> <span class="n">init_sparse_connectivity</span><span class="p">(</span><span class="s2">&quot;FixedProbability&quot;</span><span class="p">,</span> <span class="p">{</span><span class="s2">&quot;prob&quot;</span><span class="p">:</span> <span class="mf">0.1</span><span class="p">})</span>
 </pre></div>
@@ -2242,14 +2261,14 @@ <h3>Navigation</h3>
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>snippet</strong> – toeplitz connectivity init snippet, either as a string referencing
-a built in snipept (see <a class="reference internal" href="#module-pygenn.init_toeplitz_connectivity_snippets" title="pygenn.init_toeplitz_connectivity_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_toeplitz_connectivity_snippets</span></code></a>)
+a built-in snippet (see <a class="reference internal" href="#module-pygenn.init_toeplitz_connectivity_snippets" title="pygenn.init_toeplitz_connectivity_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_toeplitz_connectivity_snippets</span></code></a>)
 or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">InitToeplitzConnectivitySnippetBase</span></code>
 (for example returned by <a class="reference internal" href="#pygenn.create_toeplitz_connect_init_snippet" title="pygenn.create_toeplitz_connect_init_snippet"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_toeplitz_connect_init_snippet()</span></code></a>)</p></li>
-<li><p><strong>params</strong> – parameter values for the toeplitz connectivity init snippet (see <a href="#id52"><span class="problematic" id="id53">`Parameters`_</span></a>)</p></li>
+<li><p><strong>params</strong> – parameter values for the toeplitz connectivity init snippet (see <a href="#id54"><span class="problematic" id="id55">`Parameters`_</span></a>)</p></li>
 </ul>
 </dd>
 </dl>
-<p>For example, the built in model could be initialised to generate 2D convolutional
+<p>For example, the built-in model could be initialised to generate 2D convolutional
 connectivity with a <span class="math notranslate nohighlight">\(3 \times 3\)</span> kernel, a <span class="math notranslate nohighlight">\(64 \times 64 \times 1\)</span> input
 and a <span class="math notranslate nohighlight">\(62 \times 62 \times 1\)</span> input:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">params</span> <span class="o">=</span> <span class="p">{</span><span class="s2">&quot;conv_kh&quot;</span><span class="p">:</span> <span class="mi">3</span><span class="p">,</span> <span class="s2">&quot;conv_kw&quot;</span><span class="p">:</span> <span class="mi">3</span><span class="p">,</span>
@@ -2272,14 +2291,14 @@ <h3>Navigation</h3>
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
 <li><p><strong>snippet</strong> (<em>Union</em><em>[</em><em>InitVarSnippetBase</em><em>, </em><em>str</em><em>]</em>) – variable init snippet, either as a string referencing
-a built in snipept (see <a class="reference internal" href="#module-pygenn.init_var_snippets" title="pygenn.init_var_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_var_snippets</span></code></a>)
+a built-in snippet (see <a class="reference internal" href="#module-pygenn.init_var_snippets" title="pygenn.init_var_snippets"><code class="xref py py-mod docutils literal notranslate"><span class="pre">init_var_snippets</span></code></a>)
 or an instance of <code class="xref py py-class docutils literal notranslate"><span class="pre">InitVarSnippetBase</span></code>
 (for example returned by <a class="reference internal" href="#pygenn.create_var_init_snippet" title="pygenn.create_var_init_snippet"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_var_init_snippet()</span></code></a>)</p></li>
-<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the variable init snippet (see <a href="#id54"><span class="problematic" id="id55">`Parameters`_</span></a>)</p></li>
+<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values for the variable init snippet (see <a href="#id56"><span class="problematic" id="id57">`Parameters`_</span></a>)</p></li>
 </ul>
 </dd>
 </dl>
-<p>For example, the built in model could be initialised to initialise a variable
+<p>For example, the built-in model “Normal” could be used to initialise a variable
 by sampling from the normal distribution with a mean of 0 and a standard deviation of 1:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">init</span> <span class="o">=</span> <span class="n">init_var</span><span class="p">(</span><span class="s2">&quot;Normal&quot;</span><span class="p">,</span> <span class="p">{</span><span class="s2">&quot;mean&quot;</span><span class="p">:</span> <span class="mf">0.0</span><span class="p">,</span> <span class="s2">&quot;sd&quot;</span><span class="p">:</span> <span class="mf">1.0</span><span class="p">})</span>
 </pre></div>
@@ -2294,27 +2313,27 @@ <h3>Navigation</h3>
 <dl class="field-list simple">
 <dt class="field-odd">Parameters<span class="colon">:</span></dt>
 <dd class="field-odd"><ul class="simple">
-<li><p><strong>snippet</strong> – weight update model either as a string referencing a built in model
+<li><p><strong>snippet</strong> – weight update model either as a string referencing a built-in model
 (see <a class="reference internal" href="#module-pygenn.weight_update_models" title="pygenn.weight_update_models"><code class="xref py py-mod docutils literal notranslate"><span class="pre">weight_update_models</span></code></a>) or an instance of
 <code class="xref py py-class docutils literal notranslate"><span class="pre">WeightUpdateModelBase</span></code> (for example returned
 by <a class="reference internal" href="#pygenn.create_weight_update_model" title="pygenn.create_weight_update_model"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_weight_update_model()</span></code></a>)</p></li>
-<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values (see <a href="#id56"><span class="problematic" id="id57">`Parameters`_</span></a>)</p></li>
+<li><p><strong>params</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>int</em><em>, </em><em>float</em><em>]</em><em>]</em>) – parameter values (see <a href="#id58"><span class="problematic" id="id59">`Parameters`_</span></a>)</p></li>
 <li><p><strong>vars</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>VarInit</em><em>, </em><em>int</em><em>, </em><em>float</em><em>, </em><em>ndarray</em><em>, </em><em>Sequence</em><em>]</em><em>]</em>) – initial synaptic variable values or
-initialisers (see <a href="#id58"><span class="problematic" id="id59">`Variables`_</span></a>)</p></li>
-<li><p><strong>pre_vars</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>VarInit</em><em>, </em><em>int</em><em>, </em><em>float</em><em>, </em><em>ndarray</em><em>, </em><em>Sequence</em><em>]</em><em>]</em>) – initial presynaptic variable values or
 initialisers (see <a href="#id60"><span class="problematic" id="id61">`Variables`_</span></a>)</p></li>
+<li><p><strong>pre_vars</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>VarInit</em><em>, </em><em>int</em><em>, </em><em>float</em><em>, </em><em>ndarray</em><em>, </em><em>Sequence</em><em>]</em><em>]</em>) – initial presynaptic variable values or
+initialisers (see <a href="#id62"><span class="problematic" id="id63">`Variables`_</span></a>)</p></li>
 <li><p><strong>post_vars</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>Union</em><em>[</em><em>VarInit</em><em>, </em><em>int</em><em>, </em><em>float</em><em>, </em><em>ndarray</em><em>, </em><em>Sequence</em><em>]</em><em>]</em>) – initial postsynaptic variable values or initialisers
-(see <a href="#id62"><span class="problematic" id="id63">`Variables`_</span></a>)</p></li>
+(see <a href="#id64"><span class="problematic" id="id65">`Variables`_</span></a>)</p></li>
 <li><p><strong>pre_var_refs</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>VarReference</em><em>]</em>) – references to presynaptic neuron variables,
 typically created using <a class="reference internal" href="#pygenn.create_var_ref" title="pygenn.create_var_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_var_ref()</span></code></a>
-(see <a href="#id64"><span class="problematic" id="id65">`Variables references`_</span></a>)</p></li>
+(see <a href="#id66"><span class="problematic" id="id67">`Variables references`_</span></a>)</p></li>
 <li><p><strong>post_var_refs</strong> (<em>Dict</em><em>[</em><em>str</em><em>, </em><em>VarReference</em><em>]</em>) – references to postsynaptic neuron variables,
 typically created using <a class="reference internal" href="#pygenn.create_var_ref" title="pygenn.create_var_ref"><code class="xref py py-func docutils literal notranslate"><span class="pre">create_var_ref()</span></code></a>
-(see <a href="#id66"><span class="problematic" id="id67">`Variables references`_</span></a>)</p></li>
+(see <a href="#id68"><span class="problematic" id="id69">`Variables references`_</span></a>)</p></li>
 </ul>
 </dd>
 </dl>
-<p>For example, the built in static pulse model with
+<p>For example, the built-in static pulse model with
 constant weights could be initialised as follows:</p>
 <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="n">weight_init</span> <span class="o">=</span> <span class="n">init_weight_update</span><span class="p">(</span><span class="s2">&quot;StaticPulseConstantWeight&quot;</span><span class="p">,</span> <span class="p">{</span><span class="s2">&quot;g&quot;</span><span class="p">:</span> <span class="mf">1.0</span><span class="p">})</span>
 </pre></div>
@@ -2503,6 +2522,9 @@ <h2>Submodules<a class="headerlink" href="#submodules" title="Permalink to this
 <dd><p>Minimal custom update model for calculating tranpose</p>
 </dd></dl>
 
+</section>
+<section id="pygenn-genn-module">
+<h2>pygenn.genn module<a class="headerlink" href="#pygenn-genn-module" title="Permalink to this heading">¶</a></h2>
 </section>
 <section id="module-pygenn.genn_groups">
 <span id="pygenn-genn-groups-module"></span><h2>pygenn.genn_groups module<a class="headerlink" href="#module-pygenn.genn_groups" title="Permalink to this heading">¶</a></h2>
@@ -3515,6 +3537,7 @@ <h3><a href="../index.html">Table of Contents</a></h3>
 <li><a class="reference internal" href="#module-pygenn.current_source_models">pygenn.current_source_models module</a></li>
 <li><a class="reference internal" href="#module-pygenn.custom_connectivity_update_models">pygenn.custom_connectivity_update_models module</a></li>
 <li><a class="reference internal" href="#module-pygenn.custom_update_models">pygenn.custom_update_models module</a></li>
+<li><a class="reference internal" href="#pygenn-genn-module">pygenn.genn module</a></li>
 <li><a class="reference internal" href="#module-pygenn.genn_groups">pygenn.genn_groups module</a></li>
 <li><a class="reference internal" href="#module-pygenn.init_sparse_connectivity_snippets">pygenn.init_sparse_connectivity_snippets module</a></li>
 <li><a class="reference internal" href="#module-pygenn.init_toeplitz_connectivity_snippets">pygenn.init_toeplitz_connectivity_snippets module</a></li>
diff --git a/documentation/5/upgrading.html b/documentation/5/upgrading.html
index a3c8649d6..157fb563e 100644
--- a/documentation/5/upgrading.html
+++ b/documentation/5/upgrading.html
@@ -23,7 +23,7 @@
     <link rel="index" title="Index" href="genindex.html" />
     <link rel="search" title="Search" href="search.html" />
     <link rel="next" title="Building networks" href="building_networks.html" />
-    <link rel="prev" title="PyGeNN documentation" href="index.html" /> 
+    <link rel="prev" title="Installation" href="installation.html" /> 
   </head><body>
     <div class="related" role="navigation" aria-label="related navigation">
       <h3>Navigation</h3>
@@ -38,7 +38,7 @@ <h3>Navigation</h3>
           <a href="building_networks.html" title="Building networks"
              accesskey="N">next</a> |</li>
         <li class="right" >
-          <a href="index.html" title="PyGeNN documentation"
+          <a href="installation.html" title="Installation"
              accesskey="P">previous</a> |</li>
         <li class="nav-item nav-item-0"><a href="index.html">PyGeNN  documentation</a> &#187;</li>
         <li class="nav-item nav-item-this"><a href="">Upgrading from GeNN 4</a></li> 
@@ -68,7 +68,10 @@ <h2>Syntax changes<a class="headerlink" href="#syntax-changes" title="Permalink
 <p>In order to streamline the modelling process and reduce the number of ways to achieve the same thing we have to maintain,
 several areas of GeNN syntax have changed:</p>
 <ul>
-<li><p>single PSM code string</p></li>
+<li><p>Postsynaptic models were previously implemented with two code strings: <code class="docutils literal notranslate"><span class="pre">apply_input_code</span></code> and
+<code class="docutils literal notranslate"><span class="pre">decay_code</span></code>. This was unnecessarily complex and the order these were evaluated in wasn’t obvious.
+Therefore, these have been replaced by a single <code class="docutils literal notranslate"><span class="pre">sim_code</span></code> code string from which input can be
+delivered to the postsynaptic neuron using the <code class="docutils literal notranslate"><span class="pre">injectCurrent(x)</span></code> function.</p></li>
 <li><p><a class="reference internal" href="source/pygenn.html#pygenn.GeNNModel.add_synapse_population" title="pygenn.GeNNModel.add_synapse_population"><code class="xref py py-meth docutils literal notranslate"><span class="pre">GeNNModel.add_synapse_population()</span></code></a> was becoming rather cumbersome and this was only made worse by postsynaptic models and weight update models now taking variable references.
 To improve this:</p>
 <blockquote>
@@ -80,8 +83,13 @@ <h2>Syntax changes<a class="headerlink" href="#syntax-changes" title="Permalink
 </li>
 <li><p>The row build and diagonal build state variables in sparse/toeplitz connectivity building code were really ugly and confusing. Sparse connectivity init snippets now just let the user write whatever sort of loop they want and do the initialisation outside and toeplitz reuses the for_each_synapse structure described above to do similar.</p></li>
 <li><p><code class="docutils literal notranslate"><span class="pre">GLOBALG</span></code> and <code class="docutils literal notranslate"><span class="pre">INDIVIDUALG</span></code> confused almost all new users and were really only used with <code class="docutils literal notranslate"><span class="pre">StaticPulse</span></code> weight update models. Same functionality can be achieved with a <code class="docutils literal notranslate"><span class="pre">StaticPulseConstantWeight</span></code> version with the weight as a parameter. Then I’ve renamed all the ‘obvious’ <a class="reference internal" href="source/pygenn.html#pygenn.SynapseMatrixType" title="pygenn.SynapseMatrixType"><code class="xref py py-class docutils literal notranslate"><span class="pre">SynapseMatrixType</span></code></a> variants so you just chose <a class="reference internal" href="source/pygenn.html#pygenn.SynapseMatrixType.SPARSE" title="pygenn.SynapseMatrixType.SPARSE"><code class="xref py py-attr docutils literal notranslate"><span class="pre">SynapseMatrixType.SPARSE</span></code></a>, <a class="reference internal" href="source/pygenn.html#pygenn.SynapseMatrixType.DENSE" title="pygenn.SynapseMatrixType.DENSE"><code class="xref py py-attr docutils literal notranslate"><span class="pre">SynapseMatrixType.DENSE</span></code></a>, <a class="reference internal" href="source/pygenn.html#pygenn.SynapseMatrixType.TOEPLITZ" title="pygenn.SynapseMatrixType.TOEPLITZ"><code class="xref py py-attr docutils literal notranslate"><span class="pre">SynapseMatrixType.TOEPLITZ</span></code></a> or <a class="reference internal" href="source/pygenn.html#pygenn.SynapseMatrixType.PROCEDURAL" title="pygenn.SynapseMatrixType.PROCEDURAL"><code class="xref py py-attr docutils literal notranslate"><span class="pre">SynapseMatrixType.PROCEDURAL</span></code></a> (with <a class="reference internal" href="source/pygenn.html#pygenn.SynapseMatrixType.DENSE_PROCEDURALG" title="pygenn.SynapseMatrixType.DENSE_PROCEDURALG"><code class="xref py py-attr docutils literal notranslate"><span class="pre">SynapseMatrixType.DENSE_PROCEDURALG</span></code></a> and <a class="reference internal" href="source/pygenn.html#pygenn.SynapseMatrixType.PROCEDURAL_KERNELG" title="pygenn.SynapseMatrixType.PROCEDURAL_KERNELG"><code class="xref py py-attr docutils literal notranslate"><span class="pre">SynapseMatrixType.PROCEDURAL_KERNELG</span></code></a> for more unusual options)</p></li>
-<li><p>Extra global parameters only support the ‘pointer’ form, awaiting a PR to implement settable parameters to replace the other sort</p></li>
-<li><p>Removed implicit variable referneces</p></li>
+<li><p>Extra global parameters used to be creatable with pointer types to allow models to access
+arbitrary sized arrays and non-pointer types to create parameters that could be modified at runtime.
+The latter use case is now handled much more efficiently using the <a href="#id1"><span class="problematic" id="id2">`Dynamic parameters`_</span></a> system.</p></li>
+<li><p>Weight update and postsynaptic models could previous refer to neuron variables via _implicit_
+references. For example <code class="docutils literal notranslate"><span class="pre">$(V_pre)</span></code> would refer to a variable called <code class="docutils literal notranslate"><span class="pre">V</span></code> on the presynaptic
+neuron population. This mechanism has been replaced by adding _explicit_ references to
+weight update, postsynaptic and current source model (see <a href="#id3"><span class="problematic" id="id4">`Variables references`_</span></a>).</p></li>
 </ul>
 </section>
 </section>
@@ -106,8 +114,8 @@ <h3><a href="index.html">Table of Contents</a></h3>
   </div>
   <div>
     <h4>Previous topic</h4>
-    <p class="topless"><a href="index.html"
-                          title="previous chapter">PyGeNN documentation</a></p>
+    <p class="topless"><a href="installation.html"
+                          title="previous chapter">Installation</a></p>
   </div>
   <div>
     <h4>Next topic</h4>
@@ -148,7 +156,7 @@ <h3>Navigation</h3>
           <a href="building_networks.html" title="Building networks"
              >next</a> |</li>
         <li class="right" >
-          <a href="index.html" title="PyGeNN documentation"
+          <a href="installation.html" title="Installation"
              >previous</a> |</li>
         <li class="nav-item nav-item-0"><a href="index.html">PyGeNN  documentation</a> &#187;</li>
         <li class="nav-item nav-item-this"><a href="">Upgrading from GeNN 4</a></li> 
diff --git a/documentation/5/userproject/potjans_microcircuit.html b/documentation/5/userproject/potjans_microcircuit.html
index b427dafc3..26c66c2a1 100644
--- a/documentation/5/userproject/potjans_microcircuit.html
+++ b/documentation/5/userproject/potjans_microcircuit.html
@@ -63,8 +63,9 @@ <h3>Navigation</h3>
 The simulated spontaneous activity is asynchronous irregular and cell-type specific firing rates are in agreement
 with in vivo recordings in awake animals, including the low rate of layer 2/3 excitatory cells.
 This example can be used as follows:</p>
-<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">usage</span><span class="p">:</span> <span class="n">potjans_microcircuit</span> <span class="p">[</span><span class="o">-</span><span class="n">h</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">duration</span> <span class="n">DURATION</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">neuron</span><span class="o">-</span><span class="n">scale</span> <span class="n">NEURON_SCALE</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">connectivity</span><span class="o">-</span><span class="n">scale</span> <span class="n">CONNECTIVITY_SCALE</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">kernel</span><span class="o">-</span><span class="n">profiling</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">procedural</span><span class="o">-</span><span class="n">connectivity</span><span class="p">]</span>
-                            <span class="p">[</span><span class="o">--</span><span class="n">save</span><span class="o">-</span><span class="n">data</span><span class="p">]</span>
+<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">usage</span><span class="p">:</span> <span class="n">potjans_microcircuit</span> <span class="p">[</span><span class="o">-</span><span class="n">h</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">duration</span> <span class="n">DURATION</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">neuron</span><span class="o">-</span><span class="n">scale</span> <span class="n">NEURON_SCALE</span><span class="p">]</span>
+                            <span class="p">[</span><span class="o">--</span><span class="n">connectivity</span><span class="o">-</span><span class="n">scale</span> <span class="n">CONNECTIVITY_SCALE</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">kernel</span><span class="o">-</span><span class="n">profiling</span><span class="p">]</span>
+                            <span class="p">[</span><span class="o">--</span><span class="n">procedural</span><span class="o">-</span><span class="n">connectivity</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">save</span><span class="o">-</span><span class="n">data</span><span class="p">]</span>
 </pre></div>
 </div>
 <section id="named-arguments">
diff --git a/documentation/5/userproject/superspike_demo.html b/documentation/5/userproject/superspike_demo.html
index 0e350560c..03b0ae870 100644
--- a/documentation/5/userproject/superspike_demo.html
+++ b/documentation/5/userproject/superspike_demo.html
@@ -63,7 +63,8 @@ <h3>Navigation</h3>
 learning rule to learn the transformation between fixed spike trains of
 Poisson noise and a target spiking output (by default the Radcliffe Camera at Oxford).</p>
 <p>This example can be used as follows:</p>
-<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">usage</span><span class="p">:</span> <span class="n">superspike_demo</span> <span class="p">[</span><span class="o">-</span><span class="n">h</span><span class="p">]</span> <span class="o">--</span><span class="n">record</span><span class="o">-</span><span class="n">trial</span> <span class="p">[</span><span class="n">RECORD_TRIAL</span> <span class="o">...</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">target</span><span class="o">-</span><span class="n">file</span> <span class="n">TARGET_FILE</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">num</span><span class="o">-</span><span class="n">trials</span> <span class="n">NUM_TRIALS</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">kernel</span><span class="o">-</span><span class="n">profiling</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">save</span><span class="o">-</span><span class="n">data</span><span class="p">]</span>
+<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">usage</span><span class="p">:</span> <span class="n">superspike_demo</span> <span class="p">[</span><span class="o">-</span><span class="n">h</span><span class="p">]</span> <span class="o">--</span><span class="n">record</span><span class="o">-</span><span class="n">trial</span> <span class="p">[</span><span class="n">RECORD_TRIAL</span> <span class="o">...</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">target</span><span class="o">-</span><span class="n">file</span> <span class="n">TARGET_FILE</span><span class="p">]</span>
+                       <span class="p">[</span><span class="o">--</span><span class="n">num</span><span class="o">-</span><span class="n">trials</span> <span class="n">NUM_TRIALS</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">kernel</span><span class="o">-</span><span class="n">profiling</span><span class="p">]</span> <span class="p">[</span><span class="o">--</span><span class="n">save</span><span class="o">-</span><span class="n">data</span><span class="p">]</span>
 </pre></div>
 </div>
 <section id="named-arguments">