Metraipolis › archive › dc_832f7c58bb45
Automata from Agent Traces: Failure and Next-Step Prediction — arXiv:2608.23670
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.23670
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T06:20:32.149Z
- SHA-256 of the archived text
7f2517816f00f3f71556db64d6359a9635dad7812658aea6f58f7e23af4f8dd6- Stable URL of this page
https://metraipolis.com/documento/dc_832f7c58bb45
How to cite
Automata from Agent Traces: Failure and Next-Step Prediction — arXiv:2608.23670 — arXiv (en), https://arxiv.org/abs/2608.23670 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_832f7c58bb45 (Metraipolis archive dc_832f7c58bb45, sha256 7f2517816f00f3f71556db64d6359a9635dad7812658aea6f58f7e23af4f8dd6)
BibTeX
@misc{dc832f7c58bb45,
title = {Automata from Agent Traces: Failure and Next-Step Prediction — arXiv:2608.23670},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.23670},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_832f7c58bb45; sha256 7f2517816f00f3f71556db64d6359a9635dad7812658aea6f58f7e23af4f8dd6},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.23670. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
LLM-based agents execute multi-step tasks, but their behavioral structure remains opaque: long unstructured traces resist the safety auditing and runtime monitoring that deployment requires. Existing approaches operate per-trace or success-only, so they miss the cross-run topology that links next-step and failure prediction. To recover that shared structure, we collapse an entire trace corpus into a single, compact finite-state machine (FSM) that serves as a structural substrate for the otherwise unpredictable behavior of LLM agents. Across twelve public datasets, the FSMs are compact (7-43 states), replay held-out data at >=0.997 fitness with near-identical topology across splits, and build in milliseconds. This substrate addresses both prediction goals. For next-step prediction, FSM-state context outperforms Agent Workflow Memory on every ground-truth-matched dataset. For failure prediction, per-state behavioral features reach held-out AUROC up to 0.94, and an online monitor ranks failing runs above passing ones from a partial trace, triggering early stopping well before completion. Behavioral topology thus appears shaped more by the deployment harness than by the LLM, providing a model-agnostic structural primitive for safety auditing and runtime monitoring. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-26 in cs.CL). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2608.23670 — 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 2026-10-18T06:17:29.391Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-26?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).