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
159 results for “keyword:"monitor"” in ICLR 2027. 38 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Probing Under Distribution Shift: Learning Generalizable Distributed Representations of LLM Activations▼5.8k2.33
[PDF]Action-Conditioned Predictive Consistency for Adversarial Detection in Vision-Language-Action Models▼4.3k2.00
[PDF]Local Sensitivity in Decision Tree Ensembles with Applications to Explainability and Monitoring▼5.8k2.33
[PDF]ALDTwin – A Hybrid Physics-Based and Data-Driven Digital Twin of Atomic Layer Deposition Process for Semiconductor Innovation and Manufacturing▼1.2k1.67
[PDF]Runtime Monitoring of Distributed Backdoors in Multi-Agent LLM Systems: Local Detection, Encoding Changes, and Blocking▼5.9k2.50
[PDF]Same Probe, Different Numbers: Are Activation Probes Robust to Inference-Time Numerical Non-Determinism?▲11k2.00
[PDF]The Verbalization Gap: Stress-Testing Activation Monitoring with Natural Language Autoencoders▼1361.33
[PDF]Caught in the Act: Probes Effectively Detect Sabotage and Catch Unverbalized Deception▼1.1k3.25
[PDF]Layerwise Convergence Fingerprints for Runtime Misbehavior Detection in Large Language Models▲6.1k2.00
[PDF]Which Samples to Label First? Testing Failed Picks as Distance Anchors for Rare-Class Discovery▲3.0k2.33
[PDF]Misaligned model variants serve as an effective source of training for harm detection probes▲5103.00
[PDF]Capability Detection: Monitoring Hidden and Emergent Capabilities in Tool-Using Agents▲3.0k2.33
[PDF]Crescendo-Live: Multi-Turn Jailbreaks of Full-Duplex Speech Models in Live Calls and Their Defense with Daimonion▲3.0k2.50
[PDF]When Do Models Admit They Are Wrong? Failure Disclosure Is Unstable Under Reinforcement Learning▼1.2k1.67
[PDF]ProbeFool: Evading LLM Jailbreak Activation Monitors with Probe-Agnostic Attention Routing▼4.3k2.00
Withdrawn or desk rejected (38)
[PDF]Faithful to the Persona, Unfaithful to the Decision: Why Reasoning CoTs Cannot Report the Decisions They NarrateWithdrawn1.67
[PDF]Beyond High Prices: Auditing Information-Mediated Enforcement in Recurrent Pricing AgentsWithdrawn1.33
[PDF]Relational Window Stabilization for Repeatable Single-Window PPG Disease-Background PredictionWithdrawn1.67