Metraipolis › archive › dc_8548788707e2
Fix, Not Just Resample: Bounded Compiler-Guided Repair in Lean 4 — arXiv:2610.26942
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.26942
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-21T07:05:43.795Z
- SHA-256 of the archived text
1400db0c34b5bba9069778610a5915a467bea69fd8b00763798f3f7bc854169e- Stable URL of this page
https://metraipolis.com/documento/dc_8548788707e2
How to cite
Fix, Not Just Resample: Bounded Compiler-Guided Repair in Lean 4 — arXiv:2610.26942 — arXiv (en), https://arxiv.org/abs/2610.26942 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8548788707e2 (Metraipolis archive dc_8548788707e2, sha256 1400db0c34b5bba9069778610a5915a467bea69fd8b00763798f3f7bc854169e)
BibTeX
@misc{dc8548788707e2,
title = {Fix, Not Just Resample: Bounded Compiler-Guided Repair in Lean 4 — arXiv:2610.26942},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.26942},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8548788707e2; sha256 1400db0c34b5bba9069778610a5915a467bea69fd8b00763798f3f7bc854169e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.26942. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Automated theorem provers often rely on sampling language models, but proof assistants like Lean reject each failed step with a precise error message --- information that pure sampling discards. To overcome this, we introduce a bounded, online step-level repair module. Our search samples $K$ candidate tactics at each step and repairs a failed tactic at most $R$ times. Instead of discarding failed tactics, we feed the compiler's error message to a pretrained repair model to fix individual rejected steps in place. This approach improves the solve rate over a sampling-only baseline across all tested candidate budgets ($K \in \{1,3,5,8\}$), achieving up to a $32\%$ relative gain and solving $82$ additional theorems on a $1{,}990$-theorem subset of the LeanDojo benchmark. Repair is not free: it trades extra runtime for reachable proofs. But the trade is predictable: repair pays most when sampling is constrained; its return is front-loaded, with the first repair attempt capturing $84\text{--}96\%$ of the total gain; and the gain comes almost entirely from shallow tactic-execution failures rather than structural errors such as type mismatches. This targeted repair successfully replicates on a second benchmark (Lean Workbook), nearly doubling the solve rate with zero regressions.
Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.26942 — 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 2026-10-21T07:05:43.369Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).