Skip to main content

Chris Eliasmith

University of Waterloo, Philosophy, Faculty Member

242 Followers 1 FollowingTotal Views

Papers

Learning to select actions with spiking neurons in the basal ganglia

Abstract We expand our existing spiking neuron model of decision making in the cortex and basal g... more Abstract We expand our existing spiking neuron model of decision making in the cortex and basal ganglia to include local learning on the synaptic connections between the cortex and striatum, modulated by a dopaminergic reward signal. We then compare this model to animal data in the bandit task, which is used to test rodent learning in conditions involving forced choice under rewards. Our results indicate a good match in terms of both behavioral learning results and spike patterns in the ventral striatum.

Normalization for probabilistic inference with neurons

Abstract Recently, there have been a number of proposals regarding how biologically plausible neu... more Abstract Recently, there have been a number of proposals regarding how biologically plausible neural networks might perform probabilistic inference (Rao, Neural Computation, 16 (1): 1–38, 2004; Eliasmith and Anderson, Neural engineering: computation, representation and dynamics in neurobiological systems, 2003; Ma et al., Nature Neuroscience, 9 (11): 1432–1438, 2006; Sahani and Dayan, Neural Computation, 15 (10): 2255–2279, 2003).

An Introduction to Contemporary Theories of Content

This paper is intended to introduce advanced students of philosophy of mind to central theories a... more This paper is intended to introduce advanced students of philosophy of mind to central theories and concepts that have been used to characterize mental meaning, content, and representation. I initially provide relevant historical background. Then I discuss the ubiquitous sense/reference distinction suggested by Frege. Finally, I outline the three major kinds of contemporary theories of mental representation and discuss the difficulties with each. Much of this material has been extracted from Eliasmith (2000).

Nengo and the Neural Engineering Framework: From Spikes to Cognition

As we learn more about the neural activity underlying cognitive function, there is an increasing ... more As we learn more about the neural activity underlying cognitive function, there is an increasing demand to explicitly and quantitatively connect cognitive theories to neurological details. Bridging these levels provides benefits in both directions; aspects of the cognitive theory can predict and be constrained by neurological details, and the neurological details can in turn identify important modifications to the overall cognitive theory.

A New Perspective on Representational Problems

I argue that current flaws in the methodology of contemporary cognitive science, especially neuro... more I argue that current flaws in the methodology of contemporary cognitive science, especially neuroscience, have adversely affected philosophical theorizing about the nature of representation. To highlight these flaws, I introduce a distinction between adopting the animal's perspective and the observer's perspective when characterizing representation. I provide a discussion of each and show how the former has been unduly overlooked by cognitive scientists, including neuroscientists and philosophers.

Nengo and the Neural Engineering Framework: Connecting Cognitive Theory to Neuroscience

As we learn more about the neural activity underlying cognitive function, there is an increasing ... more As we learn more about the neural activity underlying cognitive function, there is an increasing demand to explicitly and quantitatively connect cognitive theories to neurological details. Bridging these levels provides benefits in both directions; aspects of the cognitive theory can predict and be constrained by neurological details, and the neurological details can identify important modifications to the overall cognitive theory.

Exploring the Use of Random Indexing for Retrieving Information

Abstract. This report evaluates representations for documents for the purpose of searching textua... more Abstract. This report evaluates representations for documents for the purpose of searching textual corpora. These representations include novel proposals that do and do not incorporate textual structure through the use of holographic reduced representations (HRRs) which can encode relations between words (ie noun phrases, subject-verb, object-verb, and adverb-verb).

Fine-tuning and the stability of recurrent neural networks

A central criticism of standard theoretical approaches to constructing stable, recurrent model ne... more A central criticism of standard theoretical approaches to constructing stable, recurrent model networks is that the synaptic connection weights need to be finely-tuned. This criticism is severe because proposed rules for learning these weights have been shown to have various limitations to their biological plausibility. Hence it is unlikely that such rules are used to continuously fine-tune the network in vivo. We describe a learning rule that is able to tune synaptic weights in a biologically plausible manner.

Linearly Decodable Functions From Neural Population

The population vector is a linear decoder for an ensemble of neurons, whose response properties a... more The population vector is a linear decoder for an ensemble of neurons, whose response properties are nonlinear functions of the input vector. However, previous analyses of this decoder seem to have missed the obsevation that thepopulation vector can also be used to estimate functions of the input vector. Weexplore how to use singular value decomposition to delineate the class of functionswhich are linearly decodable from a given population of noisy neural encoders.

The neural optimal control hierarchy for motor control

Abstract Our empirical, neuroscientific understanding of biological motor systems has been rapidl... more Abstract Our empirical, neuroscientific understanding of biological motor systems has been rapidly growing in recent years. However, this understanding has not been systematically mapped to a quantitative characterization of motor control based in control theory. Here, we attempt to bridge this gap by describing the neural optimal control hierarchy (NOCH), which can serve as a foundation for biologically plausible models of neural motor control.

Symbolic reasoning in spiking neurons: A model of the cortex/basal ganglia/thalamus loop

Abstract We present a model of symbol manipulation implemented using spiking neurons and closely ... more Abstract We present a model of symbol manipulation implemented using spiking neurons and closely tied to the anatomy of the cortex, basal ganglia, and thalamus. The model is a generalpurpose neural controller which plays a role analogous to a production system. Information stored in cortex is used by the basal ganglia as the basis for selecting between a set of inferences. When an inference rule is selected, it commands the thalamus to modify and transmit information between areas of the cortex.

Is the brain a quantum computer?

Abstract We argue that computation via quantum mechanical processes is irrelevant to explaining h... more Abstract We argue that computation via quantum mechanical processes is irrelevant to explaining how brains produce thought, contrary to the ongoing speculations of many theorists. First, quantum effects do not have the temporal properties required for neural information processing. Second, there are substantial physical obstacles to any organic instantiation of quantum computation.

This is a penultimate draft. To appear in P. Thagard (ed.), Philosophy of Psychology and Cognitive Science. Handbook of Philosophy of Science (Vol. 4). Amsterdam: Elsevier.

Abstract: Taking 'computational neuroscience'to denote theoretical neuroscience, I describe accou... more Abstract: Taking 'computational neuroscience'to denote theoretical neuroscience, I describe accounts of representation, neural computation, and cognitive function consistent with recent advances in this field. During this discussion I provide an overview of neural coding and survey its possible implications for more traditional theories of the meaning of mental representations.

Silicon Neurons That Compute

We use neuromorphic chips to perform arbitrary mathematical computations for the first time. Stat... more We use neuromorphic chips to perform arbitrary mathematical computations for the first time. Static and dynamic computations are realized with heterogeneous spiking silicon neurons by programming their weighted connections. Using 4K neurons with 16M feed-forward or recurrent synaptic connections, formed by 256K local arbors, we communicate a scalar stimulus, quadratically transform its value, and compute its time integral.

Knowledge and coherence

Many contemporary philosophers favor coherence theories of knowledge (Bender 1989; Bonjour 1985; ... more Many contemporary philosophers favor coherence theories of knowledge (Bender 1989; Bonjour 1985; Davidson 1986; Harman 1986; Lehrer 1990). But the nature of coherence is usually left vague, with no method provided for determining whether a belief should be accepted or rejected on the basis of its coherence or incoherence with other beliefs. S. Haack's (1993) explication of coherence relies largely on an analogy between epistemic justification and crossword puzzles.

A Brain-Machine Interface Operating with a Real-Time Spiking Neural Network Control Algorithm

Abstract:[en] Motor prostheses aim to restore function to disabled patients. Despite compelling p... more Abstract:[en] Motor prostheses aim to restore function to disabled patients. Despite compelling proof of concept systems, barriers to clinical translation remain. One challenge is to develop a low-power, fully-implantable system that dissipates only minimal power so as not to damage tissue. To this end, we implemented a Kalman-filter based decoder via a spiking neural network (SNN) and tested it in brain-machine interface (BMI) experiments with a rhesus monkey.

Spiking ratSLAM: Modeling Rat Hippocampus Place, Grid and Boarder Cells in Spiking Neural Hardware

Abstract—We will demonstrate a model of rat hippocampus place, grid and boarder cells implemented... more Abstract—We will demonstrate a model of rat hippocampus place, grid and boarder cells implemented with the SpiNNaker spiking neural hardware, configured using the Neural Engineering Framework (NEF) package Nengo, hosted on a mobile robot. These cells are used by rats for odometry, to locate landmarks and for navigation.

Brain building 101: advances in large-scale neural simulation

Physicist Richard Feynman wrote:“That which I cannot create I do not understand.” When it comes t... more Physicist Richard Feynman wrote:“That which I cannot create I do not understand.” When it comes to the brain, we are only at the very beginning of attempting to understand how it underwrites the sophisticated, subtle, and flexible behavior we see in animals. Nevertheless, there are ongoing, well-funded attempts to create one. Perhaps most famously, the Blue Brain project has built simulations of millions of neurons. 1 IBM's Cognitive Computation project, too, has recently simulated one billion neurons. 2

Reports on the 2004 AAAI fall symposia

Abstract The Association for the Advancement of Artificial Intelligence presented its 2004 Fall S... more Abstract The Association for the Advancement of Artificial Intelligence presented its 2004 Fall Symposium Series Friday through Sunday, October 22-24 at the Hyatt Regency Crystal City in Arlington, Virginia, adjacent to Washington, DC. The symposium series was preceded by a one-day AI funding seminar.

Neural representations of compositional structures: representing and manipulating vector spaces with spiking neurons

This paper re-examines the question of localist vs. distributed neural representations using a bi... more This paper re-examines the question of localist vs. distributed neural representations using a biologically realistic framework based on the central notion of neurons having a preferred direction vector. A preferred direction vector captures the general observation that neurons fire most vigorously when the stimulus lies in a particular direction in a represented vector space. This framework has been successful in capturing a wide variety of detailed neural data, although here we focus on cognitive representation.

Learning to select actions with spiking neurons in the basal ganglia

Abstract We expand our existing spiking neuron model of decision making in the cortex and basal g... more Abstract We expand our existing spiking neuron model of decision making in the cortex and basal ganglia to include local learning on the synaptic connections between the cortex and striatum, modulated by a dopaminergic reward signal. We then compare this model to animal data in the bandit task, which is used to test rodent learning in conditions involving forced choice under rewards. Our results indicate a good match in terms of both behavioral learning results and spike patterns in the ventral striatum.

Normalization for probabilistic inference with neurons

Abstract Recently, there have been a number of proposals regarding how biologically plausible neu... more Abstract Recently, there have been a number of proposals regarding how biologically plausible neural networks might perform probabilistic inference (Rao, Neural Computation, 16 (1): 1–38, 2004; Eliasmith and Anderson, Neural engineering: computation, representation and dynamics in neurobiological systems, 2003; Ma et al., Nature Neuroscience, 9 (11): 1432–1438, 2006; Sahani and Dayan, Neural Computation, 15 (10): 2255–2279, 2003).

An Introduction to Contemporary Theories of Content

This paper is intended to introduce advanced students of philosophy of mind to central theories a... more This paper is intended to introduce advanced students of philosophy of mind to central theories and concepts that have been used to characterize mental meaning, content, and representation. I initially provide relevant historical background. Then I discuss the ubiquitous sense/reference distinction suggested by Frege. Finally, I outline the three major kinds of contemporary theories of mental representation and discuss the difficulties with each. Much of this material has been extracted from Eliasmith (2000).

Nengo and the Neural Engineering Framework: From Spikes to Cognition

As we learn more about the neural activity underlying cognitive function, there is an increasing ... more As we learn more about the neural activity underlying cognitive function, there is an increasing demand to explicitly and quantitatively connect cognitive theories to neurological details. Bridging these levels provides benefits in both directions; aspects of the cognitive theory can predict and be constrained by neurological details, and the neurological details can in turn identify important modifications to the overall cognitive theory.

A New Perspective on Representational Problems

I argue that current flaws in the methodology of contemporary cognitive science, especially neuro... more I argue that current flaws in the methodology of contemporary cognitive science, especially neuroscience, have adversely affected philosophical theorizing about the nature of representation. To highlight these flaws, I introduce a distinction between adopting the animal's perspective and the observer's perspective when characterizing representation. I provide a discussion of each and show how the former has been unduly overlooked by cognitive scientists, including neuroscientists and philosophers.

Nengo and the Neural Engineering Framework: Connecting Cognitive Theory to Neuroscience

As we learn more about the neural activity underlying cognitive function, there is an increasing ... more As we learn more about the neural activity underlying cognitive function, there is an increasing demand to explicitly and quantitatively connect cognitive theories to neurological details. Bridging these levels provides benefits in both directions; aspects of the cognitive theory can predict and be constrained by neurological details, and the neurological details can identify important modifications to the overall cognitive theory.

Exploring the Use of Random Indexing for Retrieving Information

Abstract. This report evaluates representations for documents for the purpose of searching textua... more Abstract. This report evaluates representations for documents for the purpose of searching textual corpora. These representations include novel proposals that do and do not incorporate textual structure through the use of holographic reduced representations (HRRs) which can encode relations between words (ie noun phrases, subject-verb, object-verb, and adverb-verb).

Fine-tuning and the stability of recurrent neural networks

A central criticism of standard theoretical approaches to constructing stable, recurrent model ne... more A central criticism of standard theoretical approaches to constructing stable, recurrent model networks is that the synaptic connection weights need to be finely-tuned. This criticism is severe because proposed rules for learning these weights have been shown to have various limitations to their biological plausibility. Hence it is unlikely that such rules are used to continuously fine-tune the network in vivo. We describe a learning rule that is able to tune synaptic weights in a biologically plausible manner.

Linearly Decodable Functions From Neural Population

The population vector is a linear decoder for an ensemble of neurons, whose response properties a... more The population vector is a linear decoder for an ensemble of neurons, whose response properties are nonlinear functions of the input vector. However, previous analyses of this decoder seem to have missed the obsevation that thepopulation vector can also be used to estimate functions of the input vector. Weexplore how to use singular value decomposition to delineate the class of functionswhich are linearly decodable from a given population of noisy neural encoders.

The neural optimal control hierarchy for motor control

Abstract Our empirical, neuroscientific understanding of biological motor systems has been rapidl... more Abstract Our empirical, neuroscientific understanding of biological motor systems has been rapidly growing in recent years. However, this understanding has not been systematically mapped to a quantitative characterization of motor control based in control theory. Here, we attempt to bridge this gap by describing the neural optimal control hierarchy (NOCH), which can serve as a foundation for biologically plausible models of neural motor control.

Symbolic reasoning in spiking neurons: A model of the cortex/basal ganglia/thalamus loop

Abstract We present a model of symbol manipulation implemented using spiking neurons and closely ... more Abstract We present a model of symbol manipulation implemented using spiking neurons and closely tied to the anatomy of the cortex, basal ganglia, and thalamus. The model is a generalpurpose neural controller which plays a role analogous to a production system. Information stored in cortex is used by the basal ganglia as the basis for selecting between a set of inferences. When an inference rule is selected, it commands the thalamus to modify and transmit information between areas of the cortex.

Is the brain a quantum computer?

Abstract We argue that computation via quantum mechanical processes is irrelevant to explaining h... more Abstract We argue that computation via quantum mechanical processes is irrelevant to explaining how brains produce thought, contrary to the ongoing speculations of many theorists. First, quantum effects do not have the temporal properties required for neural information processing. Second, there are substantial physical obstacles to any organic instantiation of quantum computation.

This is a penultimate draft. To appear in P. Thagard (ed.), Philosophy of Psychology and Cognitive Science. Handbook of Philosophy of Science (Vol. 4). Amsterdam: Elsevier.

Abstract: Taking 'computational neuroscience'to denote theoretical neuroscience, I describe accou... more Abstract: Taking 'computational neuroscience'to denote theoretical neuroscience, I describe accounts of representation, neural computation, and cognitive function consistent with recent advances in this field. During this discussion I provide an overview of neural coding and survey its possible implications for more traditional theories of the meaning of mental representations.

Silicon Neurons That Compute

We use neuromorphic chips to perform arbitrary mathematical computations for the first time. Stat... more We use neuromorphic chips to perform arbitrary mathematical computations for the first time. Static and dynamic computations are realized with heterogeneous spiking silicon neurons by programming their weighted connections. Using 4K neurons with 16M feed-forward or recurrent synaptic connections, formed by 256K local arbors, we communicate a scalar stimulus, quadratically transform its value, and compute its time integral.

Knowledge and coherence

Many contemporary philosophers favor coherence theories of knowledge (Bender 1989; Bonjour 1985; ... more Many contemporary philosophers favor coherence theories of knowledge (Bender 1989; Bonjour 1985; Davidson 1986; Harman 1986; Lehrer 1990). But the nature of coherence is usually left vague, with no method provided for determining whether a belief should be accepted or rejected on the basis of its coherence or incoherence with other beliefs. S. Haack's (1993) explication of coherence relies largely on an analogy between epistemic justification and crossword puzzles.

A Brain-Machine Interface Operating with a Real-Time Spiking Neural Network Control Algorithm

Abstract:[en] Motor prostheses aim to restore function to disabled patients. Despite compelling p... more Abstract:[en] Motor prostheses aim to restore function to disabled patients. Despite compelling proof of concept systems, barriers to clinical translation remain. One challenge is to develop a low-power, fully-implantable system that dissipates only minimal power so as not to damage tissue. To this end, we implemented a Kalman-filter based decoder via a spiking neural network (SNN) and tested it in brain-machine interface (BMI) experiments with a rhesus monkey.

Spiking ratSLAM: Modeling Rat Hippocampus Place, Grid and Boarder Cells in Spiking Neural Hardware

Abstract—We will demonstrate a model of rat hippocampus place, grid and boarder cells implemented... more Abstract—We will demonstrate a model of rat hippocampus place, grid and boarder cells implemented with the SpiNNaker spiking neural hardware, configured using the Neural Engineering Framework (NEF) package Nengo, hosted on a mobile robot. These cells are used by rats for odometry, to locate landmarks and for navigation.

Brain building 101: advances in large-scale neural simulation

Physicist Richard Feynman wrote:“That which I cannot create I do not understand.” When it comes t... more Physicist Richard Feynman wrote:“That which I cannot create I do not understand.” When it comes to the brain, we are only at the very beginning of attempting to understand how it underwrites the sophisticated, subtle, and flexible behavior we see in animals. Nevertheless, there are ongoing, well-funded attempts to create one. Perhaps most famously, the Blue Brain project has built simulations of millions of neurons. 1 IBM's Cognitive Computation project, too, has recently simulated one billion neurons. 2

Reports on the 2004 AAAI fall symposia

Abstract The Association for the Advancement of Artificial Intelligence presented its 2004 Fall S... more Abstract The Association for the Advancement of Artificial Intelligence presented its 2004 Fall Symposium Series Friday through Sunday, October 22-24 at the Hyatt Regency Crystal City in Arlington, Virginia, adjacent to Washington, DC. The symposium series was preceded by a one-day AI funding seminar.

Neural representations of compositional structures: representing and manipulating vector spaces with spiking neurons

This paper re-examines the question of localist vs. distributed neural representations using a bi... more This paper re-examines the question of localist vs. distributed neural representations using a biologically realistic framework based on the central notion of neurons having a preferred direction vector. A preferred direction vector captures the general observation that neurons fire most vigorously when the stimulus lies in a particular direction in a represented vector space. This framework has been successful in capturing a wide variety of detailed neural data, although here we focus on cognitive representation.