Thinking Machines: Mathematical Reasoning in the Age of LLMs Journal: AI open
Andrea Asperti, Alberto Naibo, Claudio Sacerdoti Coen · 2025
Read Thinking Machines: Mathematical Reasoning in the Age of LLMs Journal: AI open by Andrea Asperti, Alberto Naibo, Claudio Sacerdoti Coen (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-00459.txt.
Introduction
Andrea Asperti and co-authors examine how language models handle mathematical reasoning in this 2025 paper. Models have grown strong at coding and at structured symbolic tasks. That success raised hopes they could also do real mathematics.
Math comes in two forms: the natural style mathematicians write, and formal code a machine can check. Progress in formal math has proven much harder than in coding. That gap, the authors argue, raises basic questions about how these models reason. They compare proof building with program writing in depth. The paper maps what would need to change for machines to prove theorems reliably. Teachers of logic may also find the survey useful in class.
Catalog record
| CREATOR | Andrea Asperti, Alberto Naibo, Claudio Sacerdoti Coen |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 120 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-00459