Policy Improvement using Language Feedback Models

Victor Zhong, Dipendra Misra, Xingdi Yuan, Marc-Alexandre Côté

Keywords: instruction following, language feedback, language grounding, learning feedback model, imitation learning

TL;DR: We train small and efficient language feedback models to identify productive behaviour in grounded instruction following environments, then imitate this behaviour to improve policy performance.

Abstract: We introduce Language Feedback Models (LFMs) that identify desirable behaviour --- actions that help achieve tasks specified in the instruction - for imitation learning in instruction following. To train LFMs, we obtain feedback from Large Language Models (LLMs) on visual trajectories verbalized to language descriptions. First, by using LFMs to identify desirable behaviour to imitate, we improve in task-completion rate over strong behavioural cloning baselines on three distinct language grounding environments (Touchdown, ScienceWorld, and ALFWorld). Second, LFMs outperform using LLMs as experts to directly predict actions, when controlling for the number of LLM output tokens. Third, LFMs generalize to unseen environments, improving task-completion rate by 3.5-12.0% through one round of adaptation. Finally, LFMs can be modified to provide human-interpretable feedback without performance loss, allowing human verification of desirable behaviour for imitation learning.

Primary Area: Natural language processing

Submission Number: 16808