Using Natural Language and Program Abstractions to Instill Human Inductive Biases in Machines

Abstract

Strong inductive biases give humans the ability to quickly learn to perform a variety of tasks. Although meta-learning is a method to endow neural networks with useful inductive biases, agents trained by meta-learning may sometimes acquire very different strategies from humans. We show that co-training these agents on predicting representations from natural language task descriptions and programs induced to generate such tasks guides them toward more human-like inductive biases. Human-generated language descriptions and program induction models that add new learned primitives both contain abstract concepts that can compress description length. Co-training on these representations result in more human-like behavior in downstream meta-reinforcement learning agents than less abstract controls (synthetic language descriptions, program induction without learned primitives), suggesting that the abstraction supported by these representations is key.

In Proceedings of the 36th Conference on Neural Information Processing Systems (NeurIPS 2022), winner of Outstanding Paper Award

Cited by in corpus (3)

• Structure in Deep Reinforcement Learning: A Survey and Open Problems
• PREDILECT: Preferences Delineated with Zero-Shot Language-based Reasoning in Reinforcement Learning
• Leveraging Systematic Knowledge of 2D Transformations