Metraipolis › archive › dc_0cd8d3d24c49
From Evidence to Action: How Tool-Using Agents Fail — arXiv:2610.07753
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.07753
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T11:21:08.545Z
- SHA-256 of the archived text
57754f4a5ae0c492c90bfca1fd8f5ce5b054536aa3a035345b0ac6ce1c2ba7bd- Stable URL of this page
https://metraipolis.com/documento/dc_0cd8d3d24c49
How to cite
From Evidence to Action: How Tool-Using Agents Fail — arXiv:2610.07753 — arXiv (en), https://arxiv.org/abs/2610.07753 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0cd8d3d24c49 (Metraipolis archive dc_0cd8d3d24c49, sha256 57754f4a5ae0c492c90bfca1fd8f5ce5b054536aa3a035345b0ac6ce1c2ba7bd)
BibTeX
@misc{dc0cd8d3d24c49,
title = {From Evidence to Action: How Tool-Using Agents Fail — arXiv:2610.07753},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.07753},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0cd8d3d24c49; sha256 57754f4a5ae0c492c90bfca1fd8f5ce5b054536aa3a035345b0ac6ce1c2ba7bd},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.07753. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Tool-using agents make consequential changes to external state, yet correct outcomes do not guarantee that their actions were supported by evidence established beforehand. We study where this evidence-to-action chain breaks as agents move from deciding whether to act to executing single actions and dependent workflows. Across ten model-harness configurations, strong static action assessment can coexist with much weaker interactive execution. Failures often begin before execution: agents stop with incomplete investigation or act before required evidence is established. Once required evidence is obtained, single-action execution is usually reliable, while multi-action workflows additionally expose unresolved prerequisites and incomplete execution. For this analysis, we introduce SafeActBench, comprising 656 cases across six operational domains and five protocols that progress from static action judgment and investigated non-action to single- and multi-action workflows. A provenance-bound Evidence Ledger and deterministic trajectory evaluator track what information was established, when actions occurred, and whether downstream dependencies were satisfied. These results show that failures arise not only from missing information, but also from how agents use established evidence when deciding and executing actions. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.07753 — 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 2026-10-07T11:15:24.761Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).