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
48 results for “keyword:"ai alignment"” in ICLR 2027. 11 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Human Difficulty Is Not Task Difficulty: What Policy Representations Know Beyond an Optimal Solver▼1.2k1.67
[PDF]Watch what LLMs do, not what they say: Preference elicitation of AI agents through game performance▼4.5k2.67
[PDF]From Behavior to Mechanism: Tracing Divergent Response Modes in Frontier Language Models▼1351.33
[PDF]Fireman against Jody: Causal Counterfactual Mirrors for Auditing Strategic Deception in Language Models▼1.2k1.67
[PDF]What Happens to the Goal? Distinct Failures of Policy Alignment in Goal Misgeneralization▲4012.67
[PDF]The Strong, weak and benign Goodhart's law: An independence-free and paradigm-agnostic formalisation▲5.4k3.00
[PDF]Don't Inoculate Everything: Stratified Inoculation Prompting Narrows Backdoor Triggers and Preserves Desired Traits▲4012.67
[PDF]When Pleasing the User Compromises Care: AI Sycophancy in Caregiver-Mediated Healthcare▼4.3k2.00
[PDF]Mind the Supersession Gap: Diagnosing Event-to-Preference Inference Failures in Language Models▼4.3k2.00
[PDF]Inoculation Adapters: Improved Selective Generalization of Capabilities with Fewer Surprising Backdoors▲3.0k2.33
[PDF]RefusalTrace: Localizing and Validating the Internal Refusal Pathway of LLMs with Sparse Autoencoder Features▲9.7k2.75
[PDF]Loyal Agents: Training LLM Agents to Protect Principal Interests Under Strategic Information Asymmetry▲3.0k2.33
[PDF]The Answer Is Not the State: Trajectory-Level Corrigibility in Diffusion Language Models▼6.2k2.67
[PDF]Defence against the dark arts of distillation: Unsupervised discovery of subliminal learning▲4012.67
Withdrawn or desk rejected (11)
[PDF]Shared Parametric Geometry and Data Provenance in Model Organisms for Emergent MisalignmentWithdrawn2.00
[PDF]Greedy alignment methods are probably (and provably) more efficient than you thinkWithdrawn1.67