Limits of PRM-Guided Tree Search for Mathematical Reasoning with LLMs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cinquin, Tristan, Pleiss, Geoff, Kristiadi, Agustinus
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917036214452224
author Cinquin, Tristan
Pleiss, Geoff
Kristiadi, Agustinus
author_facet Cinquin, Tristan
Pleiss, Geoff
Kristiadi, Agustinus
contents While chain-of-thought prompting with Best-of-N (BoN) selection has become popular for mathematical reasoning in large language models (LLMs), its linear structure fails to capture the branching and exploratory nature of complex problem-solving. In this work, we propose an adaptive algorithm to maximize process reward model (PRM) scores over the intractable action space, and investigate whether PRM-guided tree search can improve mathematical reasoning by exploring multiple partial solution paths. Across $23$ diverse mathematical problems using Qwen2.5-Math-7B-Instruct with its associated PRM as a case study, we find that: (1) PRM-guided tree search shows no statistically significant improvements over BoN despite higher costs, (2) Monte Carlo tree search and beam search outperform other PRM-guided tree search methods, (3) PRMs poorly approximate state values and their reliability degrades with reasoning depth, and (4) PRMs generalize poorly out of distribution. This underperformance stems from tree search's greater reliance on unreliable PRM scores, suggesting different reward modeling is necessary before tree search can effectively enhance mathematical reasoning in LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2510_20272
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Limits of PRM-Guided Tree Search for Mathematical Reasoning with LLMs
Cinquin, Tristan
Pleiss, Geoff
Kristiadi, Agustinus
Machine Learning
Artificial Intelligence
While chain-of-thought prompting with Best-of-N (BoN) selection has become popular for mathematical reasoning in large language models (LLMs), its linear structure fails to capture the branching and exploratory nature of complex problem-solving. In this work, we propose an adaptive algorithm to maximize process reward model (PRM) scores over the intractable action space, and investigate whether PRM-guided tree search can improve mathematical reasoning by exploring multiple partial solution paths. Across $23$ diverse mathematical problems using Qwen2.5-Math-7B-Instruct with its associated PRM as a case study, we find that: (1) PRM-guided tree search shows no statistically significant improvements over BoN despite higher costs, (2) Monte Carlo tree search and beam search outperform other PRM-guided tree search methods, (3) PRMs poorly approximate state values and their reliability degrades with reasoning depth, and (4) PRMs generalize poorly out of distribution. This underperformance stems from tree search's greater reliance on unreliable PRM scores, suggesting different reward modeling is necessary before tree search can effectively enhance mathematical reasoning in LLMs.
title Limits of PRM-Guided Tree Search for Mathematical Reasoning with LLMs
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2510.20272