Saved in:
Bibliographic Details
Main Authors: Lai, Junyu, Zhang, Jiakun, Xu, Shuo, Chen, Taolue, Wang, Zihang, Yang, Yao, Zhang, Jiarui, Cao, Chun, Xu, Jingwei
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.12031
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Recent advancements in large language models (LLMs) have sparked considerable interest in automated theorem proving and a prominent line of research integrates stepwise LLM-based provers into tree search. In this paper, we introduce a novel proof-state exploration approach for training data synthesis, designed to produce diverse tactics across a wide range of intermediate proof states, thereby facilitating effective one-shot fine-tuning of LLM as the policy model. We also propose an adaptive beam size strategy, which effectively takes advantage of our data synthesis method and achieves a trade-off between exploration and exploitation during tree search. Evaluations on the MiniF2F and ProofNet benchmarks demonstrate that our method outperforms strong baselines under the stringent Pass@1 metric, attaining an average pass rate of $60.74\%$ on MiniF2F and $21.18\%$ on ProofNet. These results underscore the impact of large-scale synthetic data in advancing automated theorem proving.