How to Build a Brain: A Neural Architecture for Biological Cognition by Chris El

Publisher Description

One goal of researchers in neuroscience, psychology, and artificial intelligence is to build theoretical models that can explain the flexibility and adaptiveness of biological systems. How to Build a Brain provides a guided exploration of a new cognitive architecture that takes biological detail seriously while addressing cognitive phenomena. The Semantic Pointer Architecture (SPA) introduced in this book provides a set of tools for constructing a widerange of biologically constrained perceptual, cognitive, and motor models. Examples of such models are provided to explain a wide range of data including single-cell recordings, neuralpopulation activity, reaction times, error rates, choice behavior, and fMRI signals. Each of the models addressed in the book introduces a major feature of biological cognition, including semantics, syntax, control, learning, and memory. These models are presented as integrated considerations of brain function, giving rise to what is currently the world's largest functional brain model. The book also compares the Semantic Pointer Architecture with the current state of theart, addressing issues of theory construction in the behavioral sciences, semantic compositionality, and scalability, among other considerations. The book concludes with a discussion of conceptualchallenges raised by this architecture, and identifies several outstanding challenges for SPA and other cognitive architectures. Along the way, the book considers neural coding, concept representation, neural dynamics, working memory, neuroanatomy, reinforcement learning, and spike-timing dependent plasticity. Eight detailed, hands-on tutorials exploiting the free Nengo neural simulation environment are also included, providing practical experience with the concepts andmodels presented throughout.

Table of Contents

1 The science of cognition1.1 The last 50 years1.2 How we got here1.3 Where we are1.4 Questions and answers1.5 Nengo: An introductionPart I. How to build a brain2 An introduction to brain building2.1 Brain parts2.2 A framework for building a brain2.2.1 Representation2.2.2 Transformation2.2.3 Dynamics2.2.4 The three principles2.3 Levels2.4 Nengo: Neural representation3 Biological cognition - Semantics3.1 The semantic pointer hypothesis3.2 What is a semantic pointer?3.3 Semantics: An overview3.4 Shallow semantics3.5 Deep semantics for perception3.6 Deep semantics for action3.7 The semantics of perception and action3.8 Nengo: Neural computations4 Biological cognition - Syntax4.1 Structured representations4.2 Binding without neurons4.3 Binding with neurons4.4 Manipulating structured representations4.5 Learning structural manipulations4.6 Clean-up memory and scaling4.7 Example: Fluid intelligence4.8 Deep semantics for cognition4.9 Nengo: Structured representations in neurons5 Biological cognition - Control5.1 The flow of information5.2 The basal ganglia5.3 Basal ganglia, cortex, and thalamus5.4 Example: Fixed sequences of actions5.5 Attention and the routing of information5.6 Example: Flexible sequences of actions5.7 Timing and control5.8 Example: The Tower of Hanoi5.9 Nengo: Question answering6 Biological cognition - Memory and learning6.1 Extending cognition through time6.2 Working memory6.3 Example: Serial list memory6.4 Biological learning6.5 Example: Learning new actions6.6 Example: Learning new syntactic manipulations6.7 Nengo: Learning7 The Semantic Pointer Architecture (SPA)7.1 A summary of the SPA7.2 A SPA unified network7.3 Tasks7.3.1 Recognition7.3.2 Copy drawing7.3.3 Reinforcement learning7.3.4 Serial working memory7.3.5 Counting7.3.6 Question answering7.3.7 Rapid variable creation7.3.8 Fluid reasoning7.3.9 Discussion7.4 A unified view: Symbols and probabilities7.5 Nengo: Advanced modeling methodsPart II. Is that how you build a brain?8 Evaluating cognitive theories8.1 Introduction8.2 Core Cognitive Criteria (CCC)8.2.1 Representational structure8.2.1.1 Systematicity8.2.1.2 Compositionality8.2.1.3 Productivity8.2.1.4 The massive binding problem8.2.2 Performance concerns8.2.2.1 Syntactic generalization8.2.2.2 Robustness8.2.2.3 Adaptability8.2.2.4 Memory8.2.2.5 Scalability8.2.3 Scientific merit8.2.3.1 Triangulation8.2.3.2 Compactness8.3 Conclusion8.4 Nengo Bonus: How to build a brain - A practical guide9 Theories of cognition9.1 The state of the art9.1.1 ACT-R9.1.2 Synchrony-based approaches9.1.3 Neural Blackboard Architecture (NBA)9.1.4 The Integrated Connectionist/Symbolic Architecture (ICS)9.1.5 Leabra9.1.6 Dynamic Field Theory (DFT)9.2 An evaluation9.2.1 Representational structure9.2.2 Performance concerns9.2.3 Scientific merit9.2.4 Summary9.3 The same...9.4 ...but different9.5 The SPA versus the SOA10 Consequences and challenges10.1 Representation10.2 Concepts10.3 Inference10.4 Dynamics10.5 Challenges10.6 ConclusionA Mathematical notation and overviewA.1 VectorsA.2 Vector spacesA.3 The dot productA.4 Basis of a vector spaceA.5 Linear transformations on vectorsA.6 Time derivatives for dynamicsB Mathematical derivations for the NEFB.1 RepresentationB.1.1 EncodingB.1.2 DecodingB.2 TransformationB.3 DynamicsC Further details on deep semantic modelsC.1 The perceptual modelC.2 The motor modelD Mathematical derivations for the SPAD.1 Binding and unbinding HRRsD.2 Learning high-level transformationsD.3 Ordinal serial encoding modelD.4 Spike-timing dependent plasticityD.5 Number of neurons for representing structureE SPA model detailsE.1 Tower of HanoiBibliographyIndex

Review

"How to Build a Brain takes on a daunting task, focusing on those parts that we think are important for memory, attention, and planning. Previous attempts at building a cognitive architecture have used symbols or connectionist networks, but Eliasmith uses spiking neurons and models specific brain regions. Categories and semantics emerge from the architecture. The way that all these moving parts work together provides insights into both the nature of cognitionand brain function."--Terrence Sejnowski, Professor and Laboratory Head, Computational Neurobiology Laboratory, Howard Hughes Medical Institute Investigator, Francis Crick Chair, Salk Institute"Eliasmith offers a unified theory of cognition that rests on the mechanism of a semantic pointer, namely, a compressed neural representation that can stand as a symbol for a more detailed semantic state or be decompressed to reproduce it, in compositional cognitive processes. Ambitious state-of-the-art modeling grounds the semantic pointer architecture in populations of spiking neurons, providing concrete neural accounts of high-level processes, includingattention, learning, memory, syntax, semantics, and reasoning. Along with offering a powerful new approach for integrating cognition and neuroscience, Eliasmith provides detailed technical accounts of hissystem, with accompanying software that will serve both students and fellow modelers well."--Lawrence W. Barsalou, Professor, Department of Psychology, Emory University

Promotional

Provides a guided exploration of a new cognitive architecture that takes biological detail seriously while addressing cognitive phenomena. Now in paperback!

Long Description

One goal of researchers in neuroscience, psychology, and artificial intelligence is to build theoretical models that can explain the flexibility and adaptiveness of biological systems. How to Build a Brain provides a guided exploration of a new cognitive architecture that takes biological detail seriously while addressing cognitive phenomena. The Semantic Pointer Architecture (SPA) introduced in this book provides a set of tools for constructing a wide rangeof biologically constrained perceptual, cognitive, and motor models. Examples of such models are provided to explain a wide range of data including single-cell recordings, neural population activity, reaction times, error rates, choice behavior, and fMRI signals. Each of the models addressed in the book introduces a major feature of biological cognition, including semantics, syntax, control, learning, and memory. These models are presented as integrated considerations of brain function, giving rise to what is currently the world's largest functional brain model. The book also compares the Semantic Pointer Architecture with the current state of the art, addressing issues of theory construction in the behavioral sciences, semantic compositionality, and scalability, among other considerations. The book concludes with a discussion of conceptual challenges raised by this architecture, and identifies several outstanding challenges for SPA and other cognitive architectures. Along the way, the book considers neural coding, concept representation, neural dynamics, working memory, neuroanatomy, reinforcement learning, and spike-timing dependent plasticity. Eight detailed, hands-on tutorials exploiting the free Nengo neural simulation environment are also included, providing practical experience with the concepts and models presented throughout.

Review Text

"How to Build a Brain takes on a daunting task, focusing on those parts that we think are important for memory, attention, and planning. Previous attempts at building a cognitive architecture have used symbols or connectionist networks, but Eliasmith uses spiking neurons and models specific brain regions. Categories and semantics emerge from the architecture. The way that all these moving parts work together provides insights into both the nature of cognitionand brain function." --Terrence Sejnowski, Professor and Laboratory Head, Computational Neurobiology Laboratory, Howard Hughes Medical Institute Investigator, Francis Crick Chair, Salk Institute"Eliasmith offers a unified theory of cognition that rests on the mechanism of a semantic pointer, namely, a compressed neural representation that can stand as a symbol for a more detailed semantic state or be decompressed to reproduce it, in compositional cognitive processes. Ambitious state-of-the-art modeling grounds the semantic pointer architecture in populations of spiking neurons, providing concrete neural accounts of high-level processes, includingattention, learning, memory, syntax, semantics, and reasoning. Along with offering a powerful new approach for integrating cognition and neuroscience, Eliasmith provides detailed technical accounts of hissystem, with accompanying software that will serve both students and fellow modelers well." --Lawrence W. Barsalou, Professor, Department of Psychology, Emory University

Review Quote

How to Build a Brain takes on a daunting task, focusing on those parts that we think are important for memory, attention, and planning. Previous attempts at building a cognitive architecture have used symbols or connectionist networks, but Eliasmith uses spiking neurons and models specific brain regions. Categories and semantics emerge from the architecture. The way that all these moving parts work together provides insights into both the nature of cognition andbrain function."

Feature

Selling point: Offers a bold new cognitive architecture based on biology called Semantic Pointer Architecture (SPA)Selling point: Presents and explains the world's largest functional brain modelSelling point: Includes practical tutorials for introducing central concepts

New Feature

1 The science of cognition 1.1 The last 50 years 1.2 How we got here 1.3 Where we are 1.4 Questions and answers 1.5 Nengo: An introduction Part I. How to build a brain 2 An introduction to brain building 2.1 Brain parts 2.2 A framework for building a brain 2.2.1 Representation 2.2.2 Transformation 2.2.3 Dynamics 2.2.4 The three principles 2.3 Levels 2.4 Nengo: Neural representation 3 Biological cognition - Semantics 3.1 The semantic pointer hypothesis 3.2 What is a semantic pointer? 3.3 Semantics: An overview 3.4 Shallow semantics 3.5 Deep semantics for perception 3.6 Deep semantics for action 3.7 The semantics of perception and action 3.8 Nengo: Neural computations 4 Biological cognition - Syntax 4.1 Structured representations 4.2 Binding without neurons 4.3 Binding with neurons 4.4 Manipulating structured representations 4.5 Learning structural manipulations 4.6 Clean-up memory and scaling 4.7 Example: Fluid intelligence 4.8 Deep semantics for cognition 4.9 Nengo: Structured representations in neurons 5 Biological cognition - Control 5.1 The flow of information 5.2 The basal ganglia 5.3 Basal ganglia, cortex, and thalamus 5.4 Example: Fixed sequences of actions 5.5 Attention and the routing of information 5.6 Example: Flexible sequences of actions 5.7 Timing and control 5.8 Example: The Tower of Hanoi 5.9 Nengo: Question answering 6 Biological cognition - Memory and learning 6.1 Extending cognition through time 6.2 Working memory 6.3 Example: Serial list memory 6.4 Biological learning 6.5 Example: Learning new actions 6.6 Example: Learning new syntactic manipulations 6.7 Nengo: Learning 7 The Semantic Pointer Architecture (SPA) 7.1 A summary of the SPA 7.2 A SPA unified network 7.3 Tasks 7.3.1 Recognition 7.3.2 Copy drawing 7.3.3 Reinforcement learning 7.3.4 Serial working memory 7.3.5 Counting 7.3.6 Question answering 7.3.7 Rapid variable creation 7.3.8 Fluid reasoning 7.3.9 Discussion 7.4 A unified view: Symbols and probabilities 7.5 Nengo: Advanced modeling methods Part II. Is that how you build a brain? 8 Evaluating cognitive theories 8.1 Introduction 8.2 Core Cognitive Criteria (CCC) 8.2.1 Representational structure 8.2.1.1 Systematicity 8.2.1.2 Compositionality 8.2.1.3 Productivity 8.2.1.4 The massive binding problem 8.2.2 Performance concerns 8.2.2.1 Syntactic generalization 8.2.2.2 Robustness 8.2.2.3 Adaptability 8.2.2.4 Memory 8.2.2.5 Scalability 8.2.3 Scientific merit 8.2.3.1 Triangulation 8.2.3.2 Compactness 8.3 Conclusion 8.4 Nengo Bonus: How to build a brain - A practical guide 9 Theories of cognition 9.1 The state of the art 9.1.1 ACT-R 9.1.2 Synchrony-based approaches 9.1.3 Neural Blackboard Architecture (NBA) 9.1.4 The Integrated Connectionist/Symbolic Architecture (ICS) 9.1.5 Leabra 9.1.6 Dynamic Field Theory (DFT) 9.2 An evaluation 9.2.1 Representational structure 9.2.2 Performance concerns 9.2.3 Scientific merit 9.2.4 Summary 9.3 The same... 9.4 ...but different 9.5 The SPA versus the SOA 10 Consequences and challenges 10.1 Representation 10.2 Concepts 10.3 Inference 10.4 Dynamics 10.5 Challenges 10.6 Conclusion A Mathematical notation and overview A.1 Vectors A.2 Vector spaces A.3 The dot product A.4 Basis of a vector space A.5 Linear transformations on vectors A.6 Time derivatives for dynamics B Mathematical derivations for the NEF B.1 Representation B.1.1 Encoding B.1.2 Decoding B.2 Transformation B.3 Dynamics C Further details on deep semantic models C.1 The perceptual model C.2 The motor model D Mathematical derivations for the SPA D.1 Binding and unbinding HRRs D.2 Learning high-level transformations D.3 Ordinal serial encoding model D.4 Spike-timing dependent plasticity D.5 Number of neurons for representing structure E SPA model details E.1 Tower of Hanoi Bibliography Index

Details