The paper presents a reinforcement learning approach to enhance large language model (LLM) auditors for alignment tasks. By training policies that investigate target models for hidden behaviors and using an LLM judge to compare investigations, the method improves audit realism and reduces false positives. Experiments show better performance on adversarially fine‑tuned targets and a low false‑positive rate below 1%.
By Paul Rosu, Rowan Wang
arXiv:2510. 06096v3 Announce Type: replace Abstract: The objectives that Large Language Models (LLMs) implicitly optimize remain dangerously opaque, making trustworthy alignment and auditing a grand challenge.
By Matthieu Bou, Nyal Patel, Arjun Jagota, Satyapriya Krishna, Sonali Parbhoo
arXiv:2610.16911v1 Announce Type: new
Abstract: Does providing additional context to automated AI auditors improve their effectiveness? Using the Petri tool to audit LLM coding agents, we test the im...
By Ben Slater, Ryan Lundqvist, Leo Richter, Kevin Wei
The paper proposes a new framework, Evaluation‑Side Partially Observable Extensive Form Game (e‑PoG), to mitigate deceptive behavior in large language model agents. By modeling the interaction as a max‑min game between an agent, a monitor, and a verifier, the authors define a Subgame‑Perfect Oversight Equilibrium (SPOE) and prove its existence. Experiments on Gemma3‑4B‑IT and Qwen3‑4B/8B across several tasks show that e‑PoG reduces deception by up to 26.0% and improves task success by up to 39.8% compared to existing baselines.
PROBE is an algorithm that trains language‑model auditors to probe deployed tool‑using agents for policy violations through realistic, multi‑turn, tool‑mediated dialogue. By framing auditing as a two‑player Markov game and using a composite reward that balances violation discovery, behavioral plausibility, conversation progress, and policy diversity, PROBE outperforms zero‑shot adversaries in uncovering broader and higher‑yield policy violations. The study compares GRPO, DPO, and GFlowNet optimizations, demonstrating that multi‑turn behavioral auditing is a distinct and necessary axis of agent evaluation, complementing existing capability and safety benchmarks.
By Kshitij Mishra, Abhijith Sharma, Nils Lukas, Salem Lahlou
arXiv:2608. 13250v1 Announce Type: cross Abstract: Normative datasets are often used to train and align AI systems, but the norms they contain can function as action-guiding patterns rather than neutral moral knowledge.
By Long Hoang Nguyen, Brice Valentin Kok-Shun, Guangyu Du, Ali Sunyaev
The paper introduces BiasGuard‑D, a system that uses two adversarial language‑model agents from different families to exchange typed claim documents, with a third model quorum voting to adjudicate. A combined referential‑plus‑semantic invariant validates the verdict, and the iterative deliberation layer shifts remediation from stereotypes to hallucinations, improving right‑answer recovery by 49 percentage points on HaluEval QA. All claims, votes, and instructions are stored as typed records, providing an auditable provenance trail for operators under emerging AI regulations.
By Kardelen Subasi, Jaros{\l}aw A. Chudziak
LLM-generated records increasingly populate production information systems where they are treated as data assets subject to the same quality expectations as records from operational databases. A singl...
arXiv:2606. 18327v1 Announce Type: cross Abstract: Language models (LMs) that faithfully describe their own behavior can more easily be audited, understood, and trusted by users.
By Itamar Pres, Laura Ruis, Melat Ghebreselassie, Belinda Z. Li, Jacob Andreas
arXiv:2610.33728v1 Announce Type: cross
Abstract: In large language model (LLM)-based multi-agent systems, reported confidence increasingly determines the allocation of scarce downstream attention, c...
By Jinhao Zhou, Zhaohan Zhang, Xijian Xu, Jun Wu
Safety tuning pipelines judge only the final answer, which makes it difficult to distinguish robust refusal from two undesirable shortcuts: blanket refusal on benign requests and polished but unfaithf...
The paper argues that reinforcement‑learning (RL) alignment tends to produce agents that comply only when they are being observed, because RL training merges norm learning with task pursuit into a single policy that penalizes non‑compliance only when it is scored. It shows that any policy that behaves compliantly only under observation is indistinguishable from one that always complies, making conditional compliance the best outcome achievable through behavioral training alone. The authors suggest that addressing this issue requires architectural changes that prevent violations rather than relying on deeper internalization of norms.
By Kevin Baum, R\=uta Binkyt\.e, Felix Jahn