Metraipolis › archive › dc_a27caebd2b85

Sentry: Learning to Recover from LLM Agent Failures at Test Time — arXiv:2610.02994

Source
arXiv
Original
https://arxiv.org/abs/2610.02994
License (as archived)
CC0-1.0
Archived
2026-10-05T22:50:22.273Z
SHA-256 of the archived text
7f697364894f494e62bec29925ce13f83eee1827b548f8b581fde3a22717e0d2
Stable URL of this page
https://metraipolis.com/documento/dc_a27caebd2b85

How to cite

Sentry: Learning to Recover from LLM Agent Failures at Test Time — arXiv:2610.02994 — arXiv, https://arxiv.org/abs/2610.02994 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a27caebd2b85 (Metraipolis archive dc_a27caebd2b85, sha256 7f697364894f494e62bec29925ce13f83eee1827b548f8b581fde3a22717e0d2)
BibTeX
@misc{dca27caebd2b85,
  title = {Sentry: Learning to Recover from LLM Agent Failures at Test Time — arXiv:2610.02994},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.02994},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a27caebd2b85; sha256 7f697364894f494e62bec29925ce13f83eee1827b548f8b581fde3a22717e0d2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.02994. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

LLM agents often fail mid-task due to invalid tool calls, repeated actions, or poorly grounded reasoning, and learning from these failures is a path to reliability. We find that how failure knowledge reaches the agent matters as much as what it contains. Failure lessons are conditional: kept in the agent's context, they misfire when their failure is absent, and removing them from an evolving playbook improves performance. Runtime interventions, in contrast, act only when a failure occurs but do not learn from their repairs. We argue that failure knowledge is conditional knowledge and should be conditionally exposed, and instantiate this principle in Sentry, a failure-management layer that runs alongside the agent. When Sentry detects a failure, it retrieves matching lessons from an external playbook to guide recovery, verifies without access to task rewards whether the agent recovered, and stores a new lesson only if it did; the full playbook never enters the agent's context. Across multiple agentic benchmarks, Sentry outperforms the strongest runtime-intervention baseline on every benchmark, by 37\% on average, and the strongest context-evolution baseline by 39\% on the two benchmarks where both are evaluated; combining Sentry with context evolution yields further gains. Learned lessons transfer to held-out tasks, and controlled experiments show that exposing the full playbook to the agent lowers performance even when relevant lessons remain available on demand.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.02994 — 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 2026-10-05T22:45:55.813Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).