Emergence of Semantic Structure from Experience Jay McClelland Stanford University.

Slides:



Advertisements
Similar presentations
Pat Langley Computational Learning Laboratory Center for the Study of Language and Information Stanford University, Stanford, California USA
Advertisements

Artificial Intelligence 12. Two Layer ANNs
ARCHITECTURES FOR ARTIFICIAL INTELLIGENCE SYSTEMS
Summer 2011 Tuesday, 8/ No supposition seems to me more natural than that there is no process in the brain correlated with associating or with.
Naïve-Bayes Classifiers Business Intelligence for Managers.
PDP: Motivation, basic approach. Cognitive psychology or “How the Mind Works”
Does the Brain Use Symbols or Distributed Representations? James L. McClelland Department of Psychology and Center for Mind, Brain, and Computation Stanford.
Knowledge ß How do we organize our knowledge? ß How do we access our knowledge? ß Do we really use categories?
Emergence in Cognitive Science: Semantic Cognition Jay McClelland Stanford University.
Emergence of Semantic Structure from Experience Jay McClelland Stanford University.
Knowing Semantic memory.
Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural.
Physical Symbol System Hypothesis
Latent Semantic Analysis (LSA). Introduction to LSA Learning Model Uses Singular Value Decomposition (SVD) to simulate human learning of word and passage.
Categorization  How do we organize our knowledge?  How do we retrieve knowledge when we need it?
Conceptual modelling. Overview - what is the aim of the article? ”We build conceptual models in our heads to solve problems in our everyday life”… ”By.
Neural Networks. Background - Neural Networks can be : Biological - Biological models Artificial - Artificial models - Desire to produce artificial systems.
Lecture 10 – Semantic Networks 1 Two questions about knowledge of the world 1.How are concepts stored? We have already considered prototype and other models.
General Knowledge Dr. Claudia J. Stanny EXP 4507 Memory & Cognition Spring 2009.
Development and Disintegration of Conceptual Knowledge: A Parallel-Distributed Processing Approach Jay McClelland Department of Psychology and Center for.
Dynamics of learning: A case study Jay McClelland Stanford University.
Using Backprop to Understand Apects of Cognitive Development PDP Class Feb 8, 2010.
Representation, Development and Disintegration of Conceptual Knowledge: A Parallel-Distributed Processing Approach James L. McClelland Department of Psychology.
Connectionism. ASSOCIATIONISM Associationism David Hume ( ) was one of the first philosophers to develop a detailed theory of mental processes.
The changing face of face research Vicki Bruce School of Psychology Newcastle University.
A Model of Object Permanence Psych 419/719 March 6, 2001.
Integrating New Findings into the Complementary Learning Systems Theory of Memory Jay McClelland, Stanford University.
The PDP Approach to Understanding the Mind and Brain J. McClelland Cognitive Core Class Lecture March 7, 2011.
The PDP Approach to Understanding the Mind and Brain Jay McClelland Stanford University January 21, 2014.
Disintegration of Conceptual Knowledge In Semantic Dementia James L. McClelland Department of Psychology and Center for Mind, Brain, and Computation Stanford.
Awakening from the Cartesian Dream: The PDP Approach to Understanding the Mind and Brain Jay McClelland Stanford University February 7, 2013.
Contrasting Approaches To Semantic Knowledge Representation and Inference Psychology 209 February 15, 2013.
Shane T. Mueller, Ph.D. Indiana University Klein Associates/ARA Rich Shiffrin Indiana University and Memory, Attention & Perception Lab REM-II: A model.
Emergence of Semantic Knowledge from Experience Jay McClelland Stanford University.
PSY 323 – COGNITION Chapter 9: Knowledge.  Categorization ◦ Process by which things are placed into groups  Concept ◦ Mental groupings of similar objects,
The Influence of Feature Type, Feature Structure and Psycholinguistic Parameters on the Naming Performance of Semantic Dementia and Alzheimer’s Patients.
Development, Disintegration, and Neural Basis of Semantic Cognition: A Parallel-Distributed Processing Approach James L. McClelland Department of Psychology.
Emergence of Semantic Structure from Experience Jay McClelland Stanford University.
Bain on Neural Networks and Connectionism Stephanie Rosenthal September 9, 2015.
Similarity and Attribution Contrasting Approaches To Semantic Knowledge Representation and Inference Jay McClelland Stanford University.
Rapid integration of new schema- consistent information in the Complementary Learning Systems Theory Jay McClelland, Stanford University.
Semantic Cognition: A Parallel Distributed Processing Approach James L. McClelland Center for the Neural Basis of Cognition and Departments of Psychology.
1 How is knowledge stored? Human knowledge comes in 2 varieties: Concepts Concepts Relations among concepts Relations among concepts So any theory of how.
The PDP Approach to Understanding the Mind and Brain Jay McClelland Stanford University January 21, 2014.
The Origins of Knowledge Debate How do people gain knowledge about the world around them? Are we born with some fundamental knowledge about concepts like.
Origins of Cognitive Abilities Jay McClelland Stanford University.
The Emergent Structure of Semantic Knowledge
The Language of Thought : Part II Joe Lau Philosophy HKU.
Emergent Semantics: Meaning and Metaphor Jay McClelland Department of Psychology and Center for Mind, Brain, and Computation Stanford University.
From NARS to a Thinking Machine Pei Wang Temple University.
Semantic Knowledge: Its Nature, its Development, and its Neural Basis James L. McClelland Department of Psychology and Center for Mind, Brain, and Computation.
Organization and Emergence of Semantic Knowledge: A Parallel-Distributed Processing Approach James L. McClelland Department of Psychology and Center for.
Development and Disintegration of Conceptual Knowledge: A Parallel-Distributed Processing Approach James L. McClelland Department of Psychology and Center.
Stages involve Discontinuous (qualitative) change Invariant sequence –Stages are never skipped.
Cognitive Modeling Cogs 4961, Cogs 6967 Psyc 4510 CSCI 4960 Mike Schoelles
Chapter 9 Knowledge. Some Questions to Consider Why is it difficult to decide if a particular object belongs to a particular category, such as “chair,”
Bayesian inference in neural networks
Psychology 209 – Winter 2017 January 31, 2017
What is cognitive psychology?
Learning linguistic structure with simple and more complex recurrent neural networks Psychology February 2, 2017.
Simple recurrent networks.
Psychology 209 – Winter 2017 Feb 28, 2017
Development and Disintegration of Conceptual Knowledge: A Parallel-Distributed Processing Approach James L. McClelland Department of Psychology and Center.
Does the Brain Use Symbols or Distributed Representations?
Emergence of Semantic Structure from Experience
Emergence of Semantics from Experience
Chap. 7 Regularization for Deep Learning (7.8~7.12 )
Artificial Intelligence 12. Two Layer ANNs
CLS, Rapid Schema Consistent Learning, and Similarity-weighted Interleaved learning Psychology 209 Feb 26, 2019.
Presentation transcript:

Emergence of Semantic Structure from Experience Jay McClelland Stanford University

The Nature of Cognition, and the Place of PDP in Cognitive Theory Many view human cognition as inherently –Structured –Systematic –Rule-governed In this framework, PDP models are seen as –Mere implementations of higher-level, rational, or ‘computational level’ models –… that don’t work as well as models that stipulate explicit rules or structures

The Alternative We argue instead that cognition (and the domains to which cognition is applied) is inherently –Quasi-regular –Semi-systematic –Context sensitive On this view, highly structured models –Are Procrustian beds into which natural cognition fits uncomfortably –Won’t capture human cognitive abilities as well as models that allow a more graded and context sensitive conception of structure

Emergent vs. Stipulated Structure Old London Midtown Manhattan

Where does structure come from? It’s built in It’s learned from experience There are constraints built in that shape what’s learned

What is the nature of the constraints? Do they specify types of structural forms? – Kemp & Tenenbaum Or are they more generic? – Lake & Tenenbaum – Most work in the PDP framework Are they constraints on the process and mechanism? Or on the space of possible outcomes? – PDP models and the K&T / L&T models rely on different sorts of constraints. – It’s far from clear that those in K&T or L&T are better.

Outline The Rumelhart model of semantic cognition, and how it relates to probabilistic models of semantic cognition. Some claims it instantiates about the nature of semantic cognition that distinguish it from K&T – Data supporting these claims, and challenging K&T Toward a fuller characterization of what the Rumelhart model learns (and one of the things it can do that sustains our interest in it). Some remarks about the so-called ‘common- sense core’.

The Rumelhart Model The Quillian Model

1.Show how learning could capture the emergence of hierarchical structure 2.Show how the model could make inferences as in the Quillian model DER’s Goals for the Model

ExperienceExperience Early Later Later Still

Start with a neutral representation on the representation units. Use backprop to adjust the representation to minimize the error.

The result is a representation similar to that of the average bird…

Use the representation to infer what this new thing can do.

Questions About the Rumelhart Model Does the model offer any advantages over other approaches? Can the mechanisms of learning and representation in the model tell us anything about Development? Effects of neuro-degeneration?

Phenomena in Development Progressive differentiation Overgeneralization of – Typical properties – Frequent names Emergent domain-specificity of representation Basic level advantage Expertise and frequency effects Conceptual reorganization

Disintegration in Semantic Dementia Loss of differentiation Overgeneralization

Distributed representations reflect different kinds of content in different areas Semantic dementia kills neurons in the temporal pole, and affects all kinds of knowledge Removal of the medial temporal lobes affects the ability to learn new semantic information, and produces a graded loss of information acquired within the last 1-30 years language The Parallel Distributed Processing Approach to Semantic Cognition

Architecture for the Organization of Semantic Memory color form motion action valance Temporal pole name Medial Temporal Lobe

Neural Networks and Probabilistic Models The Rumelhart model is learning to match the conditional probability structure of the training data: P(Attribute i = 1|Item j & Context k ) for all i,j,k The adjustments to the connection weights move them toward values than minimize a measure of the divergence between the network’s estimates of these probabilities and their values as reflected in the training data. It does so subject to strong constraints imposed by the initial connection weights and the architecture. – These constraints produce progressive differentiation, overgeneralization, etc. Depending on the structure in the training data, it can behave as though it is learning something much like one of K&T’s structure types, as well as many structures that cannot be captured exactly by any of the structures in K&T’s framework.

The Hierarchical Naïve Bayes Classifier Model (with R. Grosse and J. Glick) The world consists of things that belong to categories. Each category in turn may consist of things in several sub-categories. The features of members of each category are treated as independent –P({f i }|C j ) =  i p(f i |C j ) Knowledge of the features is acquired for the most inclusive category first. Successive layers of sub- categories emerge as evidence accumulates supporting the presence of co-occurrences violating the independence assumption. Living Things … Animals Plants Birds Fish Flowers Trees

PropertyOne-Class Model1 st class in two-class model 2 nd class in two-class model Can Grow1.0 0 Is Living1.0 0 Has Roots Has Leaves Has Branches Has Bark Has Petals Has Gills Has Scales Can Swim Can Fly Has Feathers Has Legs Has Skin Can See A One-Class and a Two-Class Naïve Bayes Classifier Model

Accounting for the network’s feature attributions with mixtures of classes at different levels of granularity Regression Beta Weight Epochs of Training Property attribution model: P(f i |item) =  k p(f i |c k ) + (1-  k )[(  j p(f i |c j ) + (1-  j )[…])

Should we replace the PDP model with the Naïve Bayes Classifier? It explains a lot of the data, and offers a succinct abstract characterization But –It only characterizes what’s learned when the data actually has hierarchical structure So it may be a useful approximate characterization in some cases, but can’t really replace the real thing.

An exploration of these ideas in the domain of mammals What is the best representation of domain content? How do people make inferences about different kinds of object properties?

Structure Extracted by a Structured Statistical Model

Predictions Similarity ratings and patterns of inference will violate the hierarchical structure Patterns of inference will vary by context

Experiments Size, predator/prey, and other properties affect similarity across birds, fish, and mammals Property inferences show clear context specificity Property inferences for blank biological properties violate predictions of K&T’s tree model for weasels, hippos, and other animals

Applying the Rumelhart model to the mammals dataset Items Contexts Behaviors Body Parts Foods Locations Movement Visual Properties

Simulation Results We look at the model’s outputs and compare these to the structure in the training data. The model captures the overall and context specific structure of the training data The model progressively extracts more and more detailed aspects of the structure as training progresses

Simulation Output

Training Data

Network Output

Progressive PCs

Learning the Structure in the Training Data Progressive sensitization to successive principal components captures learning of the mammals dataset. This subsumes the naïve Bayes classifier as a special case when there is no real cross-domain structure (as in the Quillian training corpus). So are PDP networks – and our brain’s neural networks – simply performing familiar statistical analyses?  No!

Extracting Cross- Domain Structure

v

Input Similarity Learned Similarity

A Newer Direction Exploiting knowledge sharing across domains –Lakoff: Abstract cognition is grounded in concrete reality –Boroditsky: Cognition about time is grounded in our conceptions of space Can we capture these kinds of influences through knowledge sharing across contexts? Work by Thibodeau, Glick, Sternberg & Flusberg shows that the answer is yes

Summary Distributed representations provide the substrate for learned semantic representations in the brain that develop gradually with experience and degrade gracefully with damage Succinctly stated probabilistic models can sometimes nicely approximate the structure learned, if the training data is consistent with them, but what is really learned is not likely to exactly match any such structure. The structure should be seen as a useful approximation rather than a procrustean bed into which actual knowledge must be thought of as conforming.

Comments on Tenenbaum’s Common Sense Core Project Children who supposedly have the common sense core fail simple tests they should easily pass –E.g., the A not B task Connectionist models have addressed emergence of performance in A not B and other object knowledge tasks. These models capture age-related changes as consequences of learning to maintain representations of objects that at no longer in sight. Preliminary work has been done addressing infancy studies supporting a distinction between agents and other kinds of moving objects –Information about differences between animate and inanimate objects is available in experience, and is learnable in SRN-style models that learn about objects from event structure. Initial indifference to object details would be expected in a PDP model (it’s not clear why this occurs on other approaches). Thus, as with object semantics, it is likely connectionist models will turn out to work quite well in addressing the graded development of ‘common sense knowledge’.

In Rogers and McClelland (2004) we also address: – Conceptual differentiation in prelinguistic infants. – Many of the phenomena addressed by classic work on semantic knowledge from the 1970’s: Basic level Typicality Frequency Expertise – Disintegration of conceptual knowledge in semantic dementia – How the model can be extended to capture causal properties of objects and explanations. – What properties a network must have to be sensitive to coherent covariation.