Metraipolis › archive › dc_4ad1a910ae8a

G-CARB: Graph-Localized Conformal Agent Risk Budget for Compositional Harm — arXiv:2610.05563

Source
arXiv
Original
https://arxiv.org/abs/2610.05563
License (as archived)
CC0-1.0
Archived
2026-10-06T05:26:59.738Z
SHA-256 of the archived text
c26764b6f574b808159132108feea23811c8f883fefac6a0d8bdbb94e512e8d3
Stable URL of this page
https://metraipolis.com/documento/dc_4ad1a910ae8a

How to cite

G-CARB: Graph-Localized Conformal Agent Risk Budget for Compositional Harm — arXiv:2610.05563 — arXiv, https://arxiv.org/abs/2610.05563 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4ad1a910ae8a (Metraipolis archive dc_4ad1a910ae8a, sha256 c26764b6f574b808159132108feea23811c8f883fefac6a0d8bdbb94e512e8d3)
BibTeX
@misc{dc4ad1a910ae8a,
  title = {G-CARB: Graph-Localized Conformal Agent Risk Budget for Compositional Harm — arXiv:2610.05563},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.05563},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4ad1a910ae8a; sha256 c26764b6f574b808159132108feea23811c8f883fefac6a0d8bdbb94e512e8d3},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.05563. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Small language model (SLM) agents need safety controls that track consequences across tool calls with little monitoring overhead. A private read, for example, becomes a leak when a later action sends that data outside the system. We introduce CARB (Conformal Agent Risk Budget), which calibrates when to stop an agent using a ledger of harm incurred before stopping. Under exchangeable episodes, standard conformal risk control bounds this declared loss in expectation over calibration and a future episode. G-CARB selects scorer evidence along observable dependencies from private sources to outgoing actions. The ledger still covers the entire executed history, and computing the gate score requires no additional language-model inference. On AgentDojo replay with two 14B backbones, G-CARB roughly halves scorer-input records at intermediate risk budgets while improving autonomous task completion relative to full-prefix scoring; random context of the same size achieves similar gains. Controlled examples show how retaining the relevant dependency can further avoid stopping benign work.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (stat.ML) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.05563 — 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 2026-10-06T05:24:06.043Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).