Abstract:Reasoning, the ability to logically draw conclusions from existing knowledge, is a hallmark of human. Together with perception, they constitute the two major themes of artificial intelligence. While deep learning has pushed the limit of perception beyond human-level performance, the progress in reasoning domains is way behind. One fundamental reason is that reasoning problems usually have flexible structures for both knowledge and queries, and many existing models only perform well on structures seen during training. Here we aim to push the boundary of reasoning models by devising algorithms that generalize across knowledge and query structures, as well as systems that accelerate development on structured data. This thesis consists of three parts. In Part I, we study models that can inductively generalize to unseen knowledge graphs with new entity and relation vocabularies. For new entities, we propose a framework that learns neural operators in a dynamic programming algorithm computing path representations. For relations, we construct a relation graph to capture the interactions between relations, thereby converting new relations into new entities. In Part II, we propose two solutions for generalizing across multi-step queries on knowledge graphs and text respectively. For knowledge graphs, we show that multi-step queries can be solved by multiple calls of graph neural networks and fuzzy logic operations. For text, we devise an algorithm to learn explicit knowledge as textual rules to improve large language models on multi-step queries. In Part III, we propose two systems to facilitate machine learning development on structured data. Our library treats structured data as first-class citizens and removes the barrier for developing algorithms on structured data. Our node embedding system solves the GPU memory bottleneck of embedding matrices and scales to graphs with billion nodes.

Learning Algebraic Representation for Abstract Spatial-Temporal Reasoning

Learning Algebraic Representation for Systematic Generalization in Abstract Reasoning

Abstract Spatial-Temporal Reasoning Via Probabilistic Abduction and Execution

Abstract Visual Reasoning: An Algebraic Approach for Solving Raven's Progressive Matrices

Abstract Visual Reasoning by Self-supervised Contrastive Learning

Bridging Logic and Learning: A Neural-Symbolic Approach for Enhanced Reasoning in Neural Models (ASPER)

Abstract Reasoning with Distracting Features

Stratified Rule-Aware Network for Abstract Visual Reasoning

A Feature-based Generalizable Prediction Model for Both Perceptual and Abstract Reasoning

A Cognitively-Inspired Neural Architecture for Visual Abstract Reasoning Using Contrastive Perceptual and Conceptual Processing

Learning to reason over visual objects

When Do Program-of-Thought Works for Reasoning?

Probabilistic Abduction for Visual Abstract Reasoning via Learning Rules in Vector-symbolic Architectures

Multi-Granularity Modularized Network for Abstract Visual Reasoning

Abstract Visual Reasoning Enabled by Language

Learning Robust Rule Representations for Abstract Reasoning via Internal Inferences

Neuro-symbolic Training for Reasoning over Spatial Language

Learning Representations for Reasoning: Generalizing Across Diverse Structures

Enhancing Neural Mathematical Reasoning by Abductive Combination with Symbolic Library

Meta-Reasoning: Semantics-Symbol Deconstruction for Large Language Models

Object-centric proto-symbolic behavioural reasoning from pixels