NeurIPS 2026 reviews are up: reviews, scores, rebuttals and meta-reviews for 7,877 papers. See the highest rated
acceptodds: the papers of ICLR 2027
39 results for “keyword:"oversight"” in ICLR 2027. 20 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Audit Certainty Is a Signed Lever: Why a Verifier Calibrated on a Proxy Can Backfire in Deployment▼5.8k2.33
[PDF]Capability Detection: Monitoring Hidden and Emergent Capabilities in Tool-Using Agents▲3.0k2.33
[PDF]The Irreversibility Threshold: When Error in LLM Agent Pipelines Outruns Human Correction▼1361.33
[PDF]ProbeFool: Evading LLM Jailbreak Activation Monitors with Probe-Agnostic Attention Routing▼4.3k2.00
[PDF]ControlBench: Did the Approval Actually Happen? Process-Grounded Auditing of Human Control over LLM Agents▼4.3k2.00
[PDF]ChronoCheck: Scale Improves Per-Claim Faithfulness but Shifts Time-Series Rationales Toward the Claims Models Get Wrong▼1.2k1.67
[PDF]AEGIS: Evidence-Grounded Agent Safety Verification with Diffusion Scoring and Ising-Structured Risk Inference▼4.3k2.00
[PDF]True Claims, False Answer: Spot-Check Debate Is Limited by What the Honest Side Claims▼1.2k3.00
[PDF]Behavior Cue Reasoning: Monitorable Reasoning Improves Safety and Efficiency through Oversight▲4.6k3.00
Withdrawn or desk rejected (20)
[PDF]CasePath: An Agentic, Process-First Architecture for Determining Evidence RequirementsWithdrawn1.00
[PDF]The Frontier Is Never UnderstoodDesk rejected