Compositional Generalization via Neural-Symbolic Stack Machines

Abstract

Despite achieving tremendous success, existing deep learning models have exposed limitations in compositional generalization, the capability to learn compositional rules and apply them to unseen cases in a systematic manner. To tackle this issue, we propose the Neural-Symbolic Stack Machine (NeSS). It contains a neural network to generate traces, which are then executed by a symbolic stack machine enhanced with sequence manipulation operations. NeSS combines the expressive power of neural sequence models with the recursion supported by the symbolic stack machine. Without training supervision on execution traces, NeSS achieves 100% generalization performance in four domains: the SCAN benchmark of language-driven navigation tasks, the task of few-shot learning of compositional instructions, the compositional machine translation benchmark, and context-free grammar parsing tasks.

arXiv commentPublished in NeurIPS 2020

References in corpus (21)

• Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understanding
• Neural Programmer-Interpreters
• Neural GPUs Learn Algorithms
• Inferring Algorithmic Patterns with Stack-Augmented Recurrent Nets
• The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervision
• Learning to Transduce with Unbounded Memory
• Reinforcement Learning Neural Turing Machines - Revised
• Making Neural Programming Architectures Generalize via Recursion
• Compositional generalization in a deep seq2seq model by separating syntax and semantics
• A Benchmark for Systematic Generalization in Grounded Language Understanding
• Measuring Compositional Generalization: A Comprehensive Method on Realistic Data
• Neural Logic Machines
• Automatically Composing Representation Transformations as a Means for Generalization
• CLOSURE: Assessing Systematic Generalization of CLEVR Models
• Probabilistic Neural-symbolic Models for Interpretable Visual Question Answering
• Towards Synthesizing Complex Programs from Input-Output Examples
• Learning Compositional Rules via Neural Program Synthesis
• Human few-shot learning of compositional instructions
• Learning Compositional Neural Programs with Recursive Tree Search and Planning
• Rearranging the Familiar: Testing Compositional Generalization in Recurrent Networks
• Neurally-Guided Structure Inference

Cited by in corpus (4)

• An Empirical Survey of Data Augmentation for Limited Data Learning in NLP
• Learning Algebraic Representation for Systematic Generalization in Abstract Reasoning
• Compositional Attention: Disentangling Search and Retrieval
• Can Transformers Jump Around Right in Natural Language? Assessing Performance Transfer from SCAN