{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"*Copyright (C) 2022 Intel Corporation*
\n",
"*SPDX-License-Identifier: BSD-3-Clause*
\n",
"*See: https://spdx.org/licenses/*\n",
"\n",
"---\n",
"\n",
"# Three Factor Learning with Lava \n",
"\n",
"**Motivation**: This tutorial demonstrates how Lava users can define and implement three factor learning rules using a software model of Loihi's learning engine.\n",
"\n",
"#### This tutorial assumes that you:\n",
"- have the [Lava framework installed](../../in_depth/tutorial01_installing_lava.ipynb \"Tutorial on Installing Lava\")\n",
"- are familiar with [Process interfaces in Lava](../../in_depth/tutorial02_processes.ipynb \"Tutorial on Processes\")\n",
"- are familiar with [ProcessModel implementations in Lava](../../in_depth/tutorial02_process_models.ipynb \"Tutorial on ProcessModels\")\n",
"- are familiar with how to [connect Lava Processes](../../in_depth/tutorial05_connect_processes.ipynb \"Tutorial on connecting Processes\")\n",
"- are familiar with how to [implement a custom learning rule](../../in_depth/tutorial08_stdp.ipynb \"Tutorial on STDP\")\n",
"\n",
"This tutorial demonstrates how the Lava Learning API can be used for simulation of three-factor learning rules. We first instantiate a high-level interface for the reward-modulated spike-timing dependent plasticity (R-STDP) rule defined in the Lava Process Library. We then execute a simulation of a spiking neural network in which localized, graded reward factors are mapped to specific plastic synapses, demonstrating Lava support for this new Loihi 2 learning capability. Finally, we demonstrate how users can define their own custom, three-factor learning rule interfaces and implement custom post-synaptic trace dynamics in simulation, as supported on Loihi 2 through microcoded post-synaptic neuron instructions. With these capabilities, Lava now provides support for many of the latest neuro-inspired learning algorithms under study!"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Defining three-factor learning rule interfaces in Lava\n",
"\n",
"The Lava learning API can be used to represent a wide range of local learning rules that obey Loihi's sum-of-product learning rule form. Users can define define custom, reusable learning rule interfaces atop the full Lava learning engine equation parser. Here, we import a reusable interface for reward-modulated spike-timing dependent plasticity (R-STDP) that is now part of the standard Lava Process Library.\n",
"\n",
"#### Reward-modulated Spike-Timing Dependent Plasticity (R-STDP) learning rule\n",
"\n",
"Reward-modulated STDP is a three-factor learning rule that can explain how behaviourly relevant adaptive changes in a complex network of spiking neurons could be achieved in a self-organizing manner through local synaptic plasticity. A third, reward signal modulates the outcome of the pairwise two-factor learning rule STDP. The implementation of the R-STDP described below is adapted from [Neuromodulated Spike-Timing-Dependent Plasticity](https://www.frontiersin.org/articles/10.3389/fncir.2015.00085/full \"Neuromodulated Spike-Timing-Dependent Plasticity, and Theory of Three-Factor Learning Rules\").\n",
"\n",
"Synaptic weights, $W$, are modulated as a function of the eligibility trace $E$ and reward term $R$:\n",
"\n",
"$$\\dot{W} = R \\cdot E$$\n",
"\n",
"The synaptic eligibility trace stores a temporary memory of the STDP outcome that persists through the epoch when a delayed reward signal is received. Defining the learning window of a traditional Hebbian STDP as $STDP(pre, post)$, where \"pre\" represents the pre-synaptic activity of a synapse and \"post\" represents the state of the post-synaptic neuron, the synaptic eligibility trace dynamics have the form:\n",
"\n",
"$$\\dot{E} = - \\frac{E}{\\tau_e} + STDP(pre, post)$$"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"NOTE: Learning parameters are adapted from online implemention of [Spike-Timing Dependent Plasticity (STDP)](http://www.scholarpedia.org/article/Spike-timing_dependent_plasticity \"Spike-Timing Dependent Plasticity\") and can vary based on implementation. "
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"from lava.proc.learning_rules.r_stdp_learning_rule import RewardModulatedSTDP\n",
"\n",
"R_STDP = RewardModulatedSTDP(learning_rate=1,\n",
" A_plus=-2,\n",
" A_minus=2,\n",
" pre_trace_decay_tau=10,\n",
" post_trace_decay_tau=10,\n",
" pre_trace_kernel_magnitude=16,\n",
" post_trace_kernel_magnitude=16,\n",
" eligibility_trace_decay_tau=0.5,\n",
" t_epoch=1\n",
" )"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Defining a simple learning network with localized reward signals\n",
"\n",
"We can now define a simple spiking network in which connections are modified according to our R-STDP learning rule. The core of the learning network is comprised of one pre-synaptic leaky-integrate-and-fire (LIF) neuron and two post-synaptic LIF neurons, all of which are driven by binary spiking inputs. The pre- and post-synaptic neurons are connected via a LearningDense Process, which learns weights according to the R-STDP learning rule. The binary spiking inputs drive the pre- and post- populations to affect the dynamics of the R-STDP eligibility trace.\n",
"\n",
"The network's post-synaptic population also receives localized, graded reward spikes. The post-synaptic neurons transform the graded reward spikes to localized reward traces, which act as targeted modulators to specific plastic synapses within the LearningDense process. This tutorials now includes a RSTDPLIF Process, used for post-synaptic neurons that can each compute up to two custom reward traces or \"third factor\" modulators. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"_NOTE : Though the RSTDPLIF Process can currently only be executed on CPU backend, it is modeled from Loihi 2's ability to compute custom post-traces in microcoded neuron instructions. Lava support for on-chip execution will be available soon!_"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![R_STDP_architecture](https://raw.githubusercontent.com/lava-nc/lava-docs/main/_static/images/tutorial_learning/r_stdp_tutorial_architecture.svg)"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"import numpy as np\n",
"\n",
"SELECT_TAG = \"floating_pt\""
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Initialize network parameters and weights"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"# LIF neuron parameters \n",
"du = 1\n",
"dv = 1\n",
"vth = 240\n",
"\n",
"# Pre-synaptic neuron parameters \n",
"num_neurons_pre = 1\n",
"shape_lif_pre = (num_neurons_pre, )\n",
"shape_conn_pre = (num_neurons_pre, num_neurons_pre)\n",
"\n",
"# SpikeIn -> pre-synaptic LIF connection weight \n",
"wgt_inp_pre = np.eye(num_neurons_pre) * 250\n",
"\n",
"\n",
"# Post-synaptic neuron parameters\n",
"num_neurons_post = 2\n",
"shape_lif_post = (num_neurons_post, )\n",
"shape_conn_post = (num_neurons_post, num_neurons_pre)\n",
"\n",
"# SpikeIn -> post-synaptic LIF connection weight \n",
"wgt_inp_post = np.eye(num_neurons_post) * 250\n",
"\n",
"\n",
"# Third-factor input weights\n",
"# Graded SpikeIn -> post-synaptic LIF connection weight \n",
"wgt_inp_reward = np.eye(num_neurons_post) * 100\n",
"\n",
"\n",
"# pre-LIF -> post-LIF connection initial weight (learning-enabled)\n",
"# Plastic synapse\n",
"wgt_plast_conn = np.full(shape_conn_post, 50)\n",
" "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Generate binary input and graded reward spikes"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"# Number of simulation time steps\n",
"num_steps = 200\n",
"time = list(range(1, num_steps + 1))\n",
"\n",
"# Spike times\n",
"spike_prob = 0.03\n",
"\n",
"# Create random spike rasters\n",
"np.random.seed(156)\n",
"spike_raster_pre = np.zeros((num_neurons_pre, num_steps))\n",
"np.place(spike_raster_pre, np.random.rand(num_neurons_pre, num_steps) < spike_prob, 1)\n",
"\n",
"spike_raster_post = np.zeros((num_neurons_post, num_steps))\n",
"np.place(spike_raster_post, np.random.rand(num_neurons_post, num_steps) < spike_prob, 1)\n",
"\n",
"# Create graded reward input spikes\n",
"graded_reward_spikes = np.zeros((num_neurons_post, num_steps)) \n",
"for index in range(num_steps):\n",
" if index in range(25, 50):\n",
" graded_reward_spikes[0][index] = 20\n",
" if index in range(150, 175):\n",
" graded_reward_spikes[1][index] = 40"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Initialize Network Processes"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"from lava.proc.lif.process import LIF\n",
"from lava.proc.io.source import RingBuffer as SpikeIn\n",
"from lava.proc.dense.process import LearningDense, Dense \n",
"from utils import RSTDPLIF, RSTDPLIFModel"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"# Create input devices\n",
"pattern_pre = SpikeIn(data=spike_raster_pre.astype(int))\n",
"pattern_post = SpikeIn(data=spike_raster_post.astype(int))\n",
"\n",
"# Create graded reward input device\n",
"reward_pattern_post = SpikeIn(data=graded_reward_spikes.astype(float))\n",
"\n",
"# Create input connectivity\n",
"conn_inp_pre = Dense(weights=wgt_inp_pre)\n",
"conn_inp_post = Dense(weights=wgt_inp_post)\n",
"conn_inp_reward = Dense(weights=wgt_inp_reward, num_message_bits=5)\n",
"\n",
"# Create pre-synaptic neurons\n",
"lif_pre = LIF(u=0,\n",
" v=0,\n",
" du=du,\n",
" dv=du,\n",
" bias_mant=0,\n",
" bias_exp=0,\n",
" vth=vth,\n",
" shape=shape_lif_pre,\n",
" name='lif_pre')\n",
"\n",
"# Create plastic connection\n",
"plast_conn = LearningDense(weights=wgt_plast_conn,\n",
" learning_rule=R_STDP,\n",
" name='plastic_dense')\n",
"\n",
"# Create post-synaptic neuron\n",
"lif_post = RSTDPLIF(u=0,\n",
" v=0,\n",
" du=du,\n",
" dv=du,\n",
" bias_mant=0,\n",
" bias_exp=0,\n",
" vth=vth,\n",
" shape=shape_lif_post,\n",
" learning_rule=R_STDP,\n",
" name='lif_post')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Connect Network Processes"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"# Connect network\n",
"pattern_pre.s_out.connect(conn_inp_pre.s_in)\n",
"conn_inp_pre.a_out.connect(lif_pre.a_in)\n",
"\n",
"pattern_post.s_out.connect(conn_inp_post.s_in)\n",
"conn_inp_post.a_out.connect(lif_post.a_in)\n",
"\n",
"# Reward ports\n",
"reward_pattern_post.s_out.connect(conn_inp_reward.s_in)\n",
"conn_inp_reward.a_out.connect(lif_post.a_graded_reward_in)\n",
"\n",
"lif_pre.s_out.connect(plast_conn.s_in)\n",
"plast_conn.a_out.connect(lif_post.a_in)\n",
"\n",
"# Connect reward trace callback (y2)\n",
"lif_post.s_out_y1.connect(plast_conn.s_in_y1)\n",
"lif_post.s_out_y2.connect(plast_conn.s_in_y2)\n",
"lif_post.s_out_y3.connect(plast_conn.s_in_y3)"
]
},
{
"cell_type": "markdown",
"metadata": {
"pycharm": {
"name": "#%% md\n"
}
},
"source": [
"### Create monitors to observe the weight and trace dynamics during learning"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"from lava.proc.monitor.process import Monitor\n",
"\n",
"# Create monitors\n",
"mon_pre_trace = Monitor()\n",
"mon_post_trace = Monitor()\n",
"mon_reward_trace = Monitor()\n",
"mon_pre_spikes = Monitor()\n",
"mon_post_spikes = Monitor()\n",
"mon_weight = Monitor()\n",
"mon_tag = Monitor()\n",
"mon_s_in_y2 = Monitor()\n",
"mon_y2 = Monitor()\n",
"\n",
"# Connect monitors\n",
"mon_pre_trace.probe(plast_conn.x1, num_steps)\n",
"mon_post_trace.probe(plast_conn.y1, num_steps)\n",
"mon_reward_trace.probe(lif_post.s_out_y2, num_steps)\n",
"mon_pre_spikes.probe(lif_pre.s_out, num_steps)\n",
"mon_post_spikes.probe(lif_post.s_out, num_steps)\n",
"mon_weight.probe(plast_conn.weights, num_steps)\n",
"mon_tag.probe(plast_conn.tag_1, num_steps)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Run the network\n",
"\n",
"The spiking neural network is simulated for 200 time steps using the Loihi2 simulation configuration. "
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"from lava.magma.core.run_conditions import RunSteps\n",
"from lava.magma.core.run_configs import Loihi2SimCfg"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [],
"source": [
"# Running\n",
"pattern_pre.run(condition=RunSteps(num_steps=num_steps), run_cfg=Loihi2SimCfg(select_tag=SELECT_TAG))"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [],
"source": [
"# Get data from monitors\n",
"pre_trace = mon_pre_trace.get_data()['plastic_dense']['x1']\n",
"post_trace = mon_post_trace.get_data()['plastic_dense']['y1']\n",
"reward_trace = mon_reward_trace.get_data()['lif_post']['s_out_y2']\n",
"pre_spikes = mon_pre_spikes.get_data()['lif_pre']['s_out']\n",
"post_spikes = mon_post_spikes.get_data()['lif_post']['s_out']\n",
"weights_neuron_A = mon_weight.get_data()['plastic_dense']['weights'][:, 0, 0]\n",
"weights_neuron_B = mon_weight.get_data()['plastic_dense']['weights'][:, 1, 0]\n",
"tag_neuron_A = mon_tag.get_data()['plastic_dense']['tag_1'][:, 0, 0]\n",
"tag_neuron_B = mon_tag.get_data()['plastic_dense']['tag_1'][:, 1, 0]"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"# Stopping\n",
"pattern_pre.stop()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Visualize the learning results"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"#### Plot eligibility trace dynamics"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [],
"source": [
"from utils import plot_spikes, plot_time_series, plot_time_series_subplots"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
"