Intrinsic Vicarious Conditioning for Deep Reinforcement Learning

TL;DR

This paper introduces vicarious conditioning as an intrinsic reward mechanism in deep reinforcement learning, enabling agents to learn from others without direct access to policies or reward functions, supporting low-shot and continual learning.

Contribution

It proposes a biologically-inspired vicarious conditioning framework that overcomes direct sampling limitations and demonstrates improved performance in complex environments.

Findings

Supports low-shot learning without needing demonstrator policies.

Enables longer episodes by discouraging non-descriptive terminal states.

Improves agent guidance toward desirable states in tested environments.

Abstract

Advancements in reinforcement learning have produced a variety of complex and useful intrinsic driving forces; crucially, these drivers operate under a direct conditioning paradigm. This form of conditioning limits our agents' capacity by restricting how they learn from the environment as well as from others. Off-policy or learn-by-example methods can learn from demonstrators' representations, but they require access to the demonstrating agent's policies or their reward functions. Our work overcomes this direct sampling limitation by introducing vicarious conditioning as an intrinsic reward mechanism. We draw from psychological and biological literature to provide a foundation for vicarious conditioning and use memory-based methods to implement its four steps: attention, retention, reproduction, and reinforcement. Crucially, our vicarious conditioning paradigms support low-shot learning and do not require the demonstrator agent's policy nor its reward functions. We evaluate our approach in the MiniWorld Sidewalk environment, one of the few public environments that features a non-descriptive terminal condition (no reward provided upon agent death), and extend it to Box2D's CarRacing environment. Our results across both environments demonstrate that vicarious conditioning enables longer episode lengths by discouraging the agent from non-descriptive terminal conditions and guiding the agent toward desirable states. Overall, this work emulates a cognitively-plausible learning paradigm better suited to problems such as single-life learning or continual learning.