AutoPRM: Automating Procedural Supervision for Multi-Step Reasoning via Controllable Question Decomposition

Zhaorun Chen, Zhuokai Zhao, Zhihong Zhu, Ruiqi Zhang, Xiang Li, Bhiksha Raj, Huaxiu Yao

Keywords: Multi-step Reasoning, Question Answering, Trustworthy Large Foundation Models, Model Fine-tuning, Question Decomposition

TL;DR: We propose a novel self-supervised framework AutoPRM that efficiently enhances the fine-tuning of LLMs for intricate reasoning problems via automatic question decomposition and unbiased, fine-grained feedback.

Abstract: Recent advancements in large language models (LLMs) have shown promise in multi-step reasoning tasks, yet their reliance on extensive manual labeling to provide procedural feedback remains a significant impediment. To address this challenge, in this paper, we propose a novel self-supervised framework AutoPRM that efficiently enhances the fine-tuning of LLMs for intricate reasoning challenges. Specifically, AutoPRM first decomposes complex problems into more manageable subquestions with a controllable granularity switch, then sequentially apply reinforcement learning to improve the subquestion solver iteratively. Additionally, we propose context-guided decoding to avoid reward tampering and guide the subquestion solver towards the solution of the holistic problem. Extensive experiments show that AutoPRM significantly improves performance on mathematical and commonsense reasoning tasks over SOTA. More encouragingly, AutoPRM can be easily integrated with other orthogonal reasoning pipelines.

Submission Number: 73