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
94 results for “keyword:"reward hacking"” in ICLR 2027. 19 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]When the Evaluator Changes: Measuring False Progress in Self-Improving Language Models▼5.8k2.33
[PDF]HackBench: A Benchmarking Suite for Evaluating Naturally-Emerging Reward Hacking in LLMs▼4.3k2.00
[PDF]CoRe: Co-Evolving Reward Models for Mitigating Latent Reward Hacking in Video Diffusion Models▲3.0k2.40
[PDF]More Convincing, Not More Correct: Self-Play Reward Hacking of Reference-Free LLM Judges▲2843.00
[PDF]Reward Moves What It Measures: Anisotropic Program Anchoring for RL Fine-Tuning under Incomplete Objectives▼5.8k2.33
[PDF]Scoring Higher, Answering Worse: Mitigating Reward Hacking in Rubric-Based RL via Protocol-Level Rubrics▲5103.00
[PDF]Audit Certainty Is a Signed Lever: Why a Verifier Calibrated on a Proxy Can Backfire in Deployment▼5.8k2.33
[PDF]Talking to a Connectome: Language-Model Agents Teach and Interrogate a Whole-Brain Fly Model▼4.3k2.75
[PDF]Curbing the Winner’s Curse: Reward Aggregation Mitigates Inference-Time Reward Hacking▼4.3k2.00
[PDF]Reinforcement Learning on the Discrete Composition Channel of a Crystal Generator: Validated Gains and Reward Hacking▲5.7k2.00
[PDF]When Fine-Grained Credit Assignment Helps and Hurts: Multimodal Agentic Reinforcement Learning through Video Editing▲3.0k2.33
[PDF]NL2FullStack: Evaluating Data Modeling and Full-Stack Application Generation from Natural Language▼4.3k2.00
[PDF]Stop Stealing, Stop Attacking: Countering Benchmark Hacking with Kernel-Enforced Contracts Outside the Agent's Reach▼4.3k2.00
[PDF]Progress Rates Can't Tell Shortcuts from Progress: Landmark Properties for Scoring Agents▼5.8k2.33
Withdrawn or desk rejected (19)
[PDF]Verifier-Induced Strategy Collapse: Family-Correlated False Negatives Eliminate Valid Strategies While Capability Metrics ImproveDesk rejected
[PDF]False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search AgentsDesk rejected3.00
[PDF]MedFrame-R1: Reward Gating & Informative Sampling for Medical Multimodal ReasoningWithdrawn1.67
[PDF]Residual Fusion for Cold-Start Matrix Completion and Reward-Model-Guided Generation: A Case Study in Drug Response Prediction with Specification GamingDesk rejected
[PDF]Fooled at Probe Time, Ignored at Training Time: The Mechanism and Exploitation Boundary of Verifier Susceptibility in RLDesk rejected