Summary of Putnambench: Evaluating Neural Theorem-provers on the Putnam Mathematical Competition, by George Tsoukalas et al.
PutnamBench: Evaluating Neural Theorem-Provers on the Putnam Mathematical Competition
by George Tsoukalas, Jasper Lee, John Jennings, Jimmy Xin, Michelle Ding, Michael Jennings, Amitayush Thakur, Swarat Chaudhuri
First submitted to arxiv on: 15 Jul 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computation and Language (cs.CL); Machine Learning (cs.LG); Logic in Computer Science (cs.LO); Programming Languages (cs.PL)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The abstract presents a new benchmark, PutnamBench, designed to evaluate neural theorem-provers’ ability to solve competition mathematics problems. The benchmark consists of 1692 hand-constructed formalizations of 640 theorems from the William Lowell Putnam Mathematical Competition. PutnamBench requires proficiency in various undergraduate math topics and is used to evaluate established neural and symbolic theorem-provers, which can only solve a handful of the problems. This establishes the benchmark as an open challenge for research on neural theorem-proving. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary PutnamBench is a new way to test how well computer programs can solve math problems. It has lots of math questions that were created by hand and are in different programming languages like Lean 4, Isabelle, and Coq. These questions cover many areas of math that students learn about in college. Some smart computers tried to answer these questions, but they didn’t do very well. This means there is a big challenge for people who want to make better computer programs that can solve math problems. |
