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
120 results for “keyword:"reward models"” in ICLR 2027. 25 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]RRC: Unlocking Generative Reward Models in LLM Reinforcement Learning via Ranking-Based Reward Construction▲5103.00
[PDF]Agreement Is Not Validity: A Preregistered Audit of Human-Anchored LLM-Judge Evaluation▼1.2k1.67
[PDF]The Solution Is the Supervision: Learning to Verify Reasoning without Human Step Labels▼5.8k2.33
[PDF]Do Process Reward Models Fail to Detect Errors, or Fail to Decide? A Controlled Study of Operating Points▲5.4k3.00
[PDF]Tracing Error Origins and Propagation: Dependency-Grounded Process Supervision for Mathematics and Physics▲6.1k2.00
[PDF]CO-OIPO: Coverage-Optimized Online IPO for Process Rewards under Intransitive Preferences▲5103.00
[PDF]PRISM-BENCH: Diagnosing Grounded and Faithful Multimodal Reasoning Beyond Final-Answer Accuracy
[PDF]Curbing the Winner’s Curse: Reward Aggregation Mitigates Inference-Time Reward Hacking▼4.3k2.00
[PDF]Annotate or Filter? The Redundancy–Rescue Mechanism and Phase Boundaries in LLM Fusion Pipelines▲5.4k3.00
Withdrawn or desk rejected (25)
[PDF]Pair2Point: Reliable as Pairwise, Efficient as Pointwise Rewards for High-Quality Text-to-Image GenerationWithdrawn1.67
[PDF]Trust Your Critic: Robust Reward Modeling and Reinforcement Learning for Faithful Image Editing and GenerationWithdrawn2.00
[PDF]When Learned Rewards Mislead Robotic Reinforcement Learning: Diagnosing and Calibrating Rewards under Distribution ShiftDesk rejected3.33