Metraipolis › archive › dc_ff132af1cd19
TwinCheck: Evidence-Grounded Negative-Twin Verification for Stateful Tool Agents — arXiv:2609.26911
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.26911
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T09:10:59.328Z
- SHA-256 of the archived text
fcad0c914fe0931157f38eb02e9668c267bbd750f9c96cb8123d47b610b7959f- Stable URL of this page
https://metraipolis.com/documento/dc_ff132af1cd19
How to cite
TwinCheck: Evidence-Grounded Negative-Twin Verification for Stateful Tool Agents — arXiv:2609.26911 — arXiv (en), https://arxiv.org/abs/2609.26911 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ff132af1cd19 (Metraipolis archive dc_ff132af1cd19, sha256 fcad0c914fe0931157f38eb02e9668c267bbd750f9c96cb8123d47b610b7959f)
BibTeX
@misc{dcff132af1cd19,
title = {TwinCheck: Evidence-Grounded Negative-Twin Verification for Stateful Tool Agents — arXiv:2609.26911},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.26911},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ff132af1cd19; sha256 fcad0c914fe0931157f38eb02e9668c267bbd750f9c96cb8123d47b610b7959f},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.26911. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
A single locally plausible tool call can derail an otherwise successful agent trajectory. Suspicion alone does not justify intervention, because the replacement itself can introduce the very failure verification is meant to prevent. We introduce TwinCheck, an inference-time verification policy that considers replacement only when the trace satisfies an evidence condition tied to a trace-local failure hypothesis. It constructs a trace-grounded counterfactual alternative, a negative twin, and replaces the agent's proposal only if the twin passes structural checks and the pairwise verifier prefers it in both candidate orders. For paired evaluation, exact replay holds the agent's parsed responses and actions fixed until the first accepted replacement, separating intervention effects from resampling. In the primary analysis of 159 multi-turn BFCL V4 tasks with complete exact-replay pairs, the complete policy raises task success for GPT-5.6 Sol from 45.3% to 58.5% (95% task-bootstrap CI [8.2, 18.8]), with no observed success-to-failure regressions. Together, these findings recast execution-boundary repair as a constrained comparison, making the counterfactual action itself the object of verification. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.26911 — 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 2026-10-08T08:13:26.454Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).