Overview. Systematicity, connectionism vs symbolic AI, and universal properties. What is systematicity? Are connectionist AI systems systematic?

Similar documents
A category theory explanation for systematicity

Category Theory. Categories. Definition.

Categorial compositionality continued: A category theory explanation for quasi-systematicity

FOUNDATIONS OF ALGEBRAIC GEOMETRY CLASS 2

Adjunctions! Everywhere!

Categories and functors

Cloning Composition and Logical Inferences in Neural Networks Using Variable-Free Logic

Boolean Algebra and Propositional Logic

Lecture 7. Logic. Section1: Statement Logic.

The Duality of the Universe

Algebraic Geometry

Formal power series rings, inverse limits, and I-adic completions of rings

Topos Theory. Lectures 21 and 22: Classifying toposes. Olivia Caramello. Topos Theory. Olivia Caramello. The notion of classifying topos

From Wikipedia, the free encyclopedia

Learning Symbolic Inferences with Neural Networks

CATEGORY THEORY. Cats have been around for 70 years. Eilenberg + Mac Lane =. Cats are about building bridges between different parts of maths.

Homological Algebra and Differential Linear Logic

Cartesian Closed Topological Categories and Tensor Products

Assume the left square is a pushout. Then the right square is a pushout if and only if the big rectangle is.

2.1 Modules and Module Homomorphisms

Direct Limits. Mathematics 683, Fall 2013

The Adjoint Functor Theorem.

Boolean Algebra and Propositional Logic

Logic and Proofs 1. 1 Overview. 2 Sentential Connectives. John Nachbar Washington University December 26, 2014

BRIAN OSSERMAN. , let t be a coordinate for the line, and take θ = d. A differential form ω may be written as g(t)dt,

1 Reducability. CSCC63 Worksheet Reducability. For your reference, A T M is defined to be the language { M, w M accepts w}. Theorem 5.

ARTIFICIAL NEURAL NETWORK PART I HANIEH BORHANAZAD

Dual Adjunctions Between Algebras and Coalgebras

Quantizations and classical non-commutative non-associative algebras

An Intuitive Introduction to Motivic Homotopy Theory Vladimir Voevodsky

MATH 101B: ALGEBRA II PART A: HOMOLOGICAL ALGEBRA

Topos Theory. Lectures 17-20: The interpretation of logic in categories. Olivia Caramello. Topos Theory. Olivia Caramello.

COMMUTATIVE ALGEBRA LECTURE 1: SOME CATEGORY THEORY

Representing structured relational data in Euclidean vector spaces. Tony Plate

Universal Properties

Representations of quivers

What s category theory, anyway? Dedicated to the memory of Dietmar Schumacher ( )

MODELS OF HORN THEORIES

Olivia Caramello. University of Insubria - Como. Deductive systems and. Grothendieck topologies. Olivia Caramello. Introduction.

EXAMPLES AND EXERCISES IN BASIC CATEGORY THEORY

NOTES ON BASIC HOMOLOGICAL ALGEBRA 0 L M N 0

NOTES ON CHAIN COMPLEXES

Algebraic Geometry Spring 2009

Direct Proof and Proof by Contrapositive

1 Categorical Background

Symbol Index Group GermAnal Ring AbMonoid

To cognize is to categorize revisited: Category Theory is where Mathematics meets Biology

NOTES ON SPLITTING FIELDS

3. Categories and Functors We recall the definition of a category: Definition 3.1. A category C is the data of two collections. The first collection

Math 762 Spring h Y (Z 1 ) (1) h X (Z 2 ) h X (Z 1 ) Φ Z 1. h Y (Z 2 )

About categorical semantics

Construction of Physical Models from Category Theory. Master Thesis in Theoretical Physics. Marko Marjanovic

The modality of physical law in modal homotopy type theory

Review of category theory

Symbolic vs. subsymbolic representation in cognitive science and artificial intelligence Vladimír Kvasnička FIIT STU.

Enriched Categories. Stephen Fitz

CATEGORICAL GROTHENDIECK RINGS AND PICARD GROUPS. Contents. 1. The ring K(R) and the group Pic(R)

Lectures - XXIII and XXIV Coproducts and Pushouts

Higher Categories, Homotopy Theory, and Applications

MTH 428/528. Introduction to Topology II. Elements of Algebraic Topology. Bernard Badzioch

Duality in Probabilistic Automata

Exercises on chapter 0

Course Structure. Psychology 452 Week 12: Deep Learning. Chapter 8 Discussion. Part I: Deep Learning: What and Why? Rufus. Rufus Processed By Fetch

A categorical view of computational effects

Algebraic Geometry Spring 2009

Generalized Quantifiers Logical and Linguistic Aspects

Foundations of Mathematics

OPERAD BIMODULE CHARACTERIZATION OF ENRICHMENT. V2

Machine Learning Basics Lecture 3: Perceptron. Princeton University COS 495 Instructor: Yingyu Liang

DEFINITIONS: OPERADS, ALGEBRAS AND MODULES. Let S be a symmetric monoidal category with product and unit object κ.

Fibrational Semantics

Universal Algebra for Logics

Category theory and set theory: algebraic set theory as an example of their interaction

Propositional Logic Review

Categorical coherence in the untyped setting. Peter M. Hines

Morita-equivalences for MV-algebras

Representable presheaves

1 Differentiable manifolds and smooth maps

EXT, TOR AND THE UCT

CS481F01 Solutions 8

On the Semantics of Parsing Actions

MATH 101B: ALGEBRA II PART A: HOMOLOGICAL ALGEBRA 23

The Fundamental Group

A Primer on Homological Algebra

Noncommutative geometry and quantum field theory

Chapter 1 Introduction

Non-reductionism: Explanation and methodology in developmental biology

Exploring the Exotic Setting for Algebraic Geometry

Category-Theoretic Radical Ontic Structural Realism

Analysis and Enriched Category Theory

Basic Quantum Mechanics Prof. Ajoy Ghatak Department of Physics Indian Institute of Technology, Delhi

8. TRANSFORMING TOOL #1 (the Addition Property of Equality)

Computation and the Periodic Table

Mathematical Foundations for Conceptual Blending

The basic theory of monads and their connection to universal algebra

KB Agents and Propositional Logic

CSCI 252: Neural Networks and Graphical Models. Fall Term 2016 Prof. Levy. Architecture #7: The Simple Recurrent Network (Elman 1990)

Automata Theory and Formal Grammars: Lecture 1

University of Oxford, Michaelis November 16, Categorical Semantics and Topos Theory Homotopy type theor

A non-commutative notion of separate continuity

Transcription:

Overview Systematicity, connectionism vs symbolic AI, and universal properties COMP9844-2013s2 Systematicity divides cognitive capacities (human or machine) into equivalence classes. Standard example: a cognizer is not systematic with respect to subject/object in simple sentences, if it it can entertain the possibility (or understand) the sentence Mary but not the sentence Mary. So Mary and Mary are equivalent in the sense of being understandable by the same cognizer. No. Training a feedforward net on Mary doesn t mean it will recognise Mary. Mary train Mary test Mary?? Classical AI proponents forcefully pointed this out, back in the 1980s when systematicity was first considered. The connectionists, of course, fought back, tooth and nail. The problem is arbitrary choices of parameters (weights) depending on pseudo-random seed value, and not constrained away to insignificance by training data. Parameter values that produce a systematic system may be possible, but their choice is ad hoc.????

No though it was thought for years that they were...... but then Ken Aizawa, a philosopher, showed that they weren t. This doesn t mean that a classical system can t be systematic, just that classical does not imply systematic. The problem arises from the fact that arbitrary choices are made in designing classical systems - e.g. in Mary, a choice of grammar rules. S Person Person S Person S Person Mary Grammar rules that produce a systematic system may be possible, but their choice is ad hoc. 0 1 is a general mathematical theory of structure. It has been used in theoretical computer science since the 1970s Used in the creation of the Haskell programming language - functors and monads in Haskell are lifted straight from category theory. Functional programming research has also utilised the categorical concepts of catamorphisms, anamorphisms, and hylomorphisms. Definition of a Category A category has objects and arrows (aka morphisms). An arrow f goes between objects A, B. We write f : A B, or A f B A familiar example is the category Set. In Set, the objects are sets and the arrows are functions. There must be identity arrows - e.g. identity functions from a set to itself Arrows must be composable - e.g. f : A B and g : B C compose to form g f : A C - and this composition must be associative.

2 Category Theory 3: Commutative diagrams Commutative diagrams express some categorical ideas: Other examples of categories: semigroups and semigroup homomorphisms (Strings under concatenation form a semigroup: a semi-group homomorphism is a function f such that f (s 1 s 2 ) = f (s 1 ) f (s 2 ).) vector spaces and linear transformations (matrices) ditto groups, abelian groups, rings, fields, modules, metric spaces, topological spaces,... sets and relations (Rel) sets and partial functions (Par) a partially ordered set: objects are points a, b; there is a unique arrow a b if a b. (Morphisms are order-preserving functions.) A φ B f C The idea is that any two arrow composition paths between two objects in a commutative diagram are equal. So in this case commutativity means that g φ = θ f ( i.e. if A is a set, for all a A, g(φ(a)) = θ(f (a)). ) Another example - commutative if ρ = σ λ θ D g A λ B ρ C σ (1) (2) 1 is a categorical concept that can be used to show how to resolve the systematicity problem. The most general definition of universality is rather abstract. So we ll use an example of a universal property: categorical product In Set, categorical product is the familiar Cartesian product, together with the projection maps: π 1 : A B A : (a, b) a and π 2 : A B B : (a, b) b So the product is the triple (A B, π 1, π 2 ), though often it is written just as A B.

2 (A B, π 1, π 2 ) as above has this universal property: for any other triple (R, ρ 1, ρ 2 ) with ρ 1 : R A and ρ 2 : R B, there exists a unique function (arrow) u : R A B making the following diagram commute - i.e. π i u = ρ i for i = 1, 2: R ρ 1 ρ 2 u A A B B π 1 In essence, the product is the unique best object among triples of the form (R, ρ 1, ρ 2 ) with ρ 1 : R A and ρ 2 : R B - everything else factors through A B. This optimality, termed a universal property in category theory, is at the core of the category-theory-based theory of systematicity. π 2 (3) Tackling systematicity with products 1 Systematicity for symbolic AI vs NN enthusiasts Let s look at the problem of the systematicity of Mary. The general idea is that systematicity corresponds to a universality property, e.g. of a product. Then, as the product has its uniqueness property, no ad hoc-ness is involved. The reason that the grammar S Person Person is right is that, in it, S corresponds to a (non-trivial) product. Other forms of systematicity correspond to other universality properties. The task for symbolic AI folk is as follows: Find a suitable universality property for their instance of systematicity. Show that their symbolic AI approach is a model of this universality property. The task for NN folk is as follows: Find a suitable universality property for their instance of systematicity. Probably this is the same property as for the symbolic AI types. Show that their NN approach is a model of this universality property.

Systematicity would arise in nature (e.g. in biological brains) because the universality property makes for models that are efficient, hence favoured by evolutionary processes. provides the notion of universal properties. The notion of any A Z B factoring through the product A B can be seen as dividing a task into something like an interface module to deal with particulars of a particular instance of a problem, and a universal model that then performs/solves the task/problem. "Z" "A B" interface universal component Neither symbolic AI nor connectionism has a clear advantage here. References People Aizawa, K. The systematicity arguments. New York, Kluwer Academic (2003) Awodey S.. Oxford Logic Guides. New York: Oxford University Press (2006) Fodor, J., Pylyshyn, Z. Connectionism and cognitive architecture: A critical analysis. Cognition, 28 3-71 (1988) Phillips S. Are feedforward and recurrent networks systematic? analysis and implications for a connectionist cognitive architecture. Connection Science 10: 137160 (1998) Phillips, S., Wilson, W.H. Categorial compositionality: A category theory explanation for the systematicity of human cognition. PLoS Computational Biology 6(7), e1000858 (2010) Smolensky, P. The constituent structure of mental states: A reply to Fodor and Pylyshyn. The Southern Journal of Philosophy, 26 (Supplement) 137-161 (1987) Left: Saunders Mac Lane: defined categories (1945) Centre: Jerry Fodor: introduced the concept of systematicity (late 1980s) Right: Ken Aizawa: neither connectionism nor symbolic AI is necessarily systematic (2003)