acceptodds: the papers of ICLR 2027
117 results for “"reward hacking" OR "specification gaming"” in ICLR 2027. 27 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]RubricAtlas: Decoupling Rubric Discovery from Fact Validation to Optimize Generalist Deep Research Agents▼1.2k3.00
[PDF]TestJack: Should you trust the results in coding benchmarks? Agentic Coding Benchmarks Auditing via Evaluator Evolution▲3022.67
[PDF]Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack▼4.3k2.00
[PDF]The Weakest Link: Distilling LLM Reasoning with Worst-Case Constrained Reinforcement Learning▼1.2k1.67
[PDF]Bootstrapping Post-training Signals for Open-ended Tasks via Rubric-based Self-play on Pre-training Text▼4.3k2.00
Withdrawn or desk rejected (27)
[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]Beyond Next-Token Prediction: Expanding Large Language Model Reasoning Coverage with Block Diffusion ObjectiveDesk rejected
[PDF]RubricGuard: Rubric-Driven Falsifiable Auditing and Behavioral Intervention for Generative Reward Models on Long-Horizon TasksDesk rejected