Learning from Mistakes via Cooperative Study Assistant for Large Language Models
This addresses the limitation of self-feedback in LLMs for improving generation accuracy, though it appears incremental as it builds on existing cooperative agent frameworks.
The paper tackles the problem of inaccurate self-feedback in large language models (LLMs) by proposing SALAM, a cooperative framework with an auxiliary agent that helps the main LLM learn from mistakes, resulting in accuracy improvements of up to 6.6 on BBH and 12.6 on BBQ.
Large language models (LLMs) have demonstrated their potential to refine their generation based on their own feedback. However, the feedback from LLM itself is often inaccurate, thereby limiting its benefits. In this paper, we propose Study Assistant for Large LAnguage Model (SALAM), a novel framework with an auxiliary agent to assist the main LLM in learning from mistakes through interactive cooperation. In the gathering phase, the student assistant agent probes the main LLM, analyzes its errors, and collects the interaction in a mistake memory. During the examination phase, the study assistant provides guidelines by retrieving relevant cases to help the main LLM anticipate and avoid similar errors. We first investigate the effectiveness of a general study assistant and then customize it to provide LLM-specific guidance through imitation learning from successful guidance experiences. Our experiments on three LLMs using two challenging frameworks demonstrate that SALAM can significantly boost LLMs by an accuracy margin of up to 6.6 on BBH and 12.6 on BBQ.