Latent Reward: LLM-Empowered Credit Assignment in Episodic Reinforcement Learning

TL;DR

This paper introduces LaRe, a novel framework that leverages Large Language Models to improve credit assignment in episodic reinforcement learning by using a multi-dimensional latent reward concept for better interpretability and reward redistribution.

Contribution

It proposes a symbolic-based decision-making framework that utilizes LLM-generated semantic code and latent reward self-verification to enhance credit assignment in RL tasks.

Findings

LaRe achieves superior temporal credit assignment compared to SOTA methods.

It effectively allocates contributions among multiple agents.

Outperforms policies trained with ground truth rewards in certain tasks.

Abstract

Reinforcement learning (RL) often encounters delayed and sparse feedback in real-world applications, even with only episodic rewards. Previous approaches have made some progress in reward redistribution for credit assignment but still face challenges, including training difficulties due to redundancy and ambiguous attributions stemming from overlooking the multifaceted nature of mission performance evaluation. Hopefully, Large Language Model (LLM) encompasses fruitful decision-making knowledge and provides a plausible tool for reward redistribution. Even so, deploying LLM in this case is non-trivial due to the misalignment between linguistic knowledge and the symbolic form requirement, together with inherent randomness and hallucinations in inference. To tackle these issues, we introduce LaRe, a novel LLM-empowered symbolic-based decision-making framework, to improve credit assignment. Key to LaRe is the concept of the Latent Reward, which works as a multi-dimensional performance evaluation, enabling more interpretable goal attainment from various perspectives and facilitating more effective reward redistribution. We examine that semantically generated code from LLM can bridge linguistic knowledge and symbolic latent rewards, as it is executable for symbolic objects. Meanwhile, we design latent reward self-verification to increase the stability and reliability of LLM inference. Theoretically, reward-irrelevant redundancy elimination in the latent reward benefits RL performance from more accurate reward estimation. Extensive experimental results witness that LaRe (i) achieves superior temporal credit assignment to SOTA methods, (ii) excels in allocating contributions among multiple agents, and (iii) outperforms policies trained with ground truth rewards for certain tasks.