Metraipolis › archive › dc_8a7374155c91
RetroChessBench: A Benchmark for Reconstructing and Proving Chess Histories — arXiv:2610.39341
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.39341
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-31T06:13:11.586Z
- SHA-256 of the archived text
0e540de72163746c3b9c031d5841051cfe2c07f4c174a65bbf5a7917b9824a22- Stable URL of this page
https://metraipolis.com/documento/dc_8a7374155c91
How to cite
RetroChessBench: A Benchmark for Reconstructing and Proving Chess Histories — arXiv:2610.39341 — arXiv (en), https://arxiv.org/abs/2610.39341 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8a7374155c91 (Metraipolis archive dc_8a7374155c91, sha256 0e540de72163746c3b9c031d5841051cfe2c07f4c174a65bbf5a7917b9824a22)
BibTeX
@misc{dc8a7374155c91,
title = {RetroChessBench: A Benchmark for Reconstructing and Proving Chess Histories — arXiv:2610.39341},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.39341},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8a7374155c91; sha256 0e540de72163746c3b9c031d5841051cfe2c07f4c174a65bbf5a7917b9824a22},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.39341. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Frontier language models such as GPT 6 Astra have contributed to research mathematics, demonstrating substantial reasoning capabilities. Yet serious reasoning failures persist even in tasks governed by explicit rules. We demonstrate this in retrograde chess which is a playground for pure reasoning, where solvers reconstruct legal histories and distinguish possible events from forced ones. We introduce RetroChessBench, a 12-task benchmark covering exact-length constructions, unique endings, and exhaustive answer sets, with references, rubrics, and a terminal-reward environment. In two trials per task and condition, GPT 6 Astra High fails to produce the required final answer in 22/24 attempts without tools and 18/24 with code execution; mean terminal RL rewards are 14.75% and 27.19%, respectively. The examples identify faulty deductions: failed responses reject legal histories and treat temporary obstructions as permanent. Online retrograde archives often provide puzzles and final move sequences without detailed derivations. We hypothesize that limited exposure to these missing explanations in web-derived training data contributes to the observed reasoning failures. This questions how far frontier mathematical reasoning really generalizes beyond familiar solution patterns. Claims of general intelligence require evidence that models can transfer learnt principles to areas with few worked examples. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.39341 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-31T06:07:08.708Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).