An Evaluation Benchmark for Autoformalization in Lean4
This provides a benchmark for researchers working on autoformalization in mathematics, but it is incremental as it focuses on evaluation rather than new methods.
The paper tackles the problem of evaluating autoformalization capabilities of Large Language Models (LLMs) in the Lean4 mathematical programming language, introducing a novel benchmark and testing models like GPT-3.5, GPT-4, and Gemini Pro, with results showing they still have limitations in complex mathematics.
Large Language Models (LLMs) hold the potential to revolutionize autoformalization. The introduction of Lean4, a mathematical programming language, presents an unprecedented opportunity to rigorously assess the autoformalization capabilities of LLMs. This paper introduces a novel evaluation benchmark designed for Lean4, applying it to test the abilities of state-of-the-art LLMs, including GPT-3.5, GPT-4, and Gemini Pro. Our comprehensive analysis reveals that, despite recent advancements, these LLMs still exhibit limitations in autoformalization, particularly in more complex areas of mathematics. These findings underscore the need for further development in LLMs to fully harness their potential in scientific research and development. This study not only benchmarks current LLM capabilities but also sets the stage for future enhancements in autoformalization.