Saved in:
Bibliographic Details
Main Authors: Han, Sangjun, Hur, Taeil, Hur, Youngmi, Lee, Kathy Sangkyung, Lee, Myungyoon, Lim, Hyojae
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.03321
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929719899848704
author Han, Sangjun
Hur, Taeil
Hur, Youngmi
Lee, Kathy Sangkyung
Lee, Myungyoon
Lim, Hyojae
author_facet Han, Sangjun
Hur, Taeil
Hur, Youngmi
Lee, Kathy Sangkyung
Lee, Myungyoon
Lim, Hyojae
contents The challenge of formal proof generation has a rich history, but with modern techniques, we may finally be at the stage of making actual progress in real-life mathematical problems. This paper explores the integration of ChatGPT and basic searching techniques to simplify generating formal proofs, with a particular focus on the miniF2F dataset. We demonstrate how combining a large language model like ChatGPT with a formal language such as Lean, which has the added advantage of being verifiable, enhances the efficiency and accessibility of formal proof generation. Despite its simplicity, our best-performing Lean-based model surpasses all known benchmarks with a 31.15% pass rate. We extend our experiments to include other datasets and employ alternative language models, showcasing our models' comparable performance in diverse settings and allowing for a more nuanced analysis of our results. Our findings offer insights into AI-assisted formal proof generation, suggesting a promising direction for future research in formal mathematical proof.
format Preprint
id arxiv_https___arxiv_org_abs_2502_03321
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Simplifying Formal Proof-Generating Models with ChatGPT and Basic Searching Techniques
Han, Sangjun
Hur, Taeil
Hur, Youngmi
Lee, Kathy Sangkyung
Lee, Myungyoon
Lim, Hyojae
Logic in Computer Science
Artificial Intelligence
The challenge of formal proof generation has a rich history, but with modern techniques, we may finally be at the stage of making actual progress in real-life mathematical problems. This paper explores the integration of ChatGPT and basic searching techniques to simplify generating formal proofs, with a particular focus on the miniF2F dataset. We demonstrate how combining a large language model like ChatGPT with a formal language such as Lean, which has the added advantage of being verifiable, enhances the efficiency and accessibility of formal proof generation. Despite its simplicity, our best-performing Lean-based model surpasses all known benchmarks with a 31.15% pass rate. We extend our experiments to include other datasets and employ alternative language models, showcasing our models' comparable performance in diverse settings and allowing for a more nuanced analysis of our results. Our findings offer insights into AI-assisted formal proof generation, suggesting a promising direction for future research in formal mathematical proof.
title Simplifying Formal Proof-Generating Models with ChatGPT and Basic Searching Techniques
topic Logic in Computer Science
Artificial Intelligence
url https://arxiv.org/abs/2502.03321