Metraipolis › archive › dc_b4372dbd16d8
Claim-Level Reliability Assessment for Efficient Test-Time Reasoning — arXiv:2608.11994
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.11994
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T05:15:32.652Z
- SHA-256 of the archived text
db6bea351fddd53b0a6311ae18a5da11951a642876d849d1bc49682a421a3210- Stable URL of this page
https://metraipolis.com/documento/dc_b4372dbd16d8
How to cite
Claim-Level Reliability Assessment for Efficient Test-Time Reasoning — arXiv:2608.11994 — arXiv (en), https://arxiv.org/abs/2608.11994 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b4372dbd16d8 (Metraipolis archive dc_b4372dbd16d8, sha256 db6bea351fddd53b0a6311ae18a5da11951a642876d849d1bc49682a421a3210)
BibTeX
@misc{dcb4372dbd16d8,
title = {Claim-Level Reliability Assessment for Efficient Test-Time Reasoning — arXiv:2608.11994},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.11994},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b4372dbd16d8; sha256 db6bea351fddd53b0a6311ae18a5da11951a642876d849d1bc49682a421a3210},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.11994. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We propose claim-level falsification as a principle for test-time scaling and instantiate it through Claim-Level Reliability Assessment (CLR), a training-free framework that reallocates test-time compute from additional solution sampling to targeted verification. Since whole-trace evaluation often obscures decisive errors due to signal dilution from routine tokens, CLR condenses each reasoning trace into a compact set of decision-critical claims, thereby isolating its logical anchors. Furthermore, recognizing the inherent difficulty of generating entirely correct solutions under fixed model capabilities, CLR shifts the focus to semantic falsification. This approach exploits a fundamental asymmetry between solution construction and claim refutation. Constructing a valid solution requires a flawless reasoning path, whereas refuting an incorrect claim requires identifying only a single decisive flaw. This targeted search for negative evidence systematically compresses the survival space of high-confidence incorrect traces, effectively suppressing erroneous consensus via nonlinear reliability scoring. Across four LLMs and four reasoning benchmarks under matched budgets, CLR generally improves upon pass@1 and self-consistency. On GPT-OSS-20B/CMIMC25, for instance, CLR exceeds pass@1 by 27.15 percentage-points and raises self-consistency accuracy from 77.50\% to 82.19\% with 37.0\% fewer tokens. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-13 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.11994 — 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 2026-10-19T05:08:46.650Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-13?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).