Metraipolis › archive › dc_5934dfd4c24f

Simulated Control Evals — arXiv:2610.31593

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.31593
License (as archived): CC0-1.0
Archived: 2026-10-25T20:54:33.051Z
SHA-256 of the archived text: b6c08c6ebc7e2455709cc6f85658d9b3abf715bc8f8bd85fbe97bef647507a05
Stable URL of this page: https://metraipolis.com/documento/dc_5934dfd4c24f

How to cite

Simulated Control Evals — arXiv:2610.31593 — arXiv (en), https://arxiv.org/abs/2610.31593 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5934dfd4c24f (Metraipolis archive dc_5934dfd4c24f, sha256 b6c08c6ebc7e2455709cc6f85658d9b3abf715bc8f8bd85fbe97bef647507a05)

BibTeX

@misc{dc5934dfd4c24f,
  title = {Simulated Control Evals — arXiv:2610.31593},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.31593},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5934dfd4c24f; sha256 b6c08c6ebc7e2455709cc6f85658d9b3abf715bc8f8bd85fbe97bef647507a05},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.31593. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Control evaluations measure whether safeguards can prevent an AI model from causing undesired outcomes even when it is intentionally trying to do so. These evaluations typically use purpose-built control settings and extrapolate their results to a model's deployment. However, the effectiveness of safeguards in deployment can depend on features of environments and tasks that control settings do not represent. To broaden where control evaluations can be run, we introduce simulated control evaluations (SCEs), which use a language model to simulate tool responses in place of an environment. We evaluate SCEs on LinuxArena, where execution provides ground-truth outcomes. We compare SCE prediction errors with those from a separate control setting, BashArena, and with variation in control metrics within LinuxArena. For independently sampled evaluation and deployment task sets, SCEs predict safety with a root mean squared error of 3.6 percentage points across five monitors at a 5% audit budget, compared with 10.1 for BashArena. SCEs closely reproduce distributions of attack strategies, attacker reasoning, and monitor reasoning, with over tenfold lower KL divergence from execution than BashArena. SCEs explain 78-85% of the between-environment variance in attack success, monitor scores, and safety. This accuracy is sample efficient: SCEs predict safety with less error than BashArena using fewer than 5% as many task-pair draws as the deployment task set. Strategies discovered through simulation also improve monitor performance on held-out LinuxArena attacks, outperforming BashArena and approximately matching improvement from executed strategies. These results suggest that deployment-specific SCEs can be more predictive than control evaluations executed in separate control settings.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.31593 — 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 2026-10-25T20:47:56.947Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).