FIMO: A Challenge Formal Dataset for Automated Theorem Proving
This addresses the challenge of automated theorem proving at high-level mathematics for researchers, but it is incremental as it focuses on dataset creation rather than a new solving method.
The authors introduced FIMO, a dataset of 149 formal mathematical problem statements from International Mathematical Olympiad (IMO) Shortlisted Problems, designed for automated theorem proving in Lean, and found through GPT-4 experiments that current methods have significant limitations in achieving satisfactory IMO-level results.
We present FIMO, an innovative dataset comprising formal mathematical problem statements sourced from the International Mathematical Olympiad (IMO) Shortlisted Problems. Designed to facilitate advanced automated theorem proving at the IMO level, FIMO is currently tailored for the Lean formal language. It comprises 149 formal problem statements, accompanied by both informal problem descriptions and their corresponding LaTeX-based informal proofs. Through initial experiments involving GPT-4, our findings underscore the existing limitations in current methodologies, indicating a substantial journey ahead before achieving satisfactory IMO-level automated theorem proving outcomes.