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
3,168 results for “keyword:"reinforcement learnin"” in ICLR 2027. 643 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]MAMR-GRPO: Reinforcement Learning with Execution-Structured Credit for Multi-Agent LLMs▼1611.50
[PDF]From-Base RLVR for Verify-Only Judges: A Rule-Based Reward Turns Base Models into Trajectory Auditors▲11.00
[PDF]Critical Decision Capacity: Bottleneck vs. Substitutable Degradation under Action Availability▼1.2k1.67
[PDF]Where, What, and When: Structured Intra-Response Advantage Redistribution for LLM Reasoning▼1361.33
[PDF]EPIC: Rethinking Trust Regions for Autoregressive LLM Policies via Exact-KL Prefix Importance-ratio Clipping▼1.2k1.67
[PDF]In Their Own Words: Reasoning Traces Tailored for Small Models Make Them Better Reasoners▲2843.00
[PDF]Rewarding the Graph Behind the Chain: Topology-Aware Process Supervision for RL and Reasoning Distillation▲3922.67
[PDF]VLNAPO: Gradient-Guided Harness Engineering with Key-Token Supervision for Full-Game VLNA Policy Optimization in OpenRA▼1.2k1.67
[PDF]SERPO: Self-Evolving Rubric Policy Optimization for Open-Ended Test-Time Reinforcement Learning▼4.3k2.00
[PDF]Pressure Is Not Enough: A Causal Test of Whether Optimization Induces Encoded Chain-of-Thought▼1.2k1.67
[PDF]Stabilizing Independent Learning in General-Sum Self-Play via Coupling-Adaptive Regularization▲133.33
[PDF]Prioritized Rollouts for Efficient World Model-based Vision-Language-Action Policy Optimization▲3.0k2.33
[PDF]Trust Labels Should Calibrate, Not Decide: Diagnosing and Reducing Source-Label Shortcuts in RL-Trained Fact-Checkers▲3.0k2.33
[PDF]Physics-Reinforced Distribution Matching Distillation for Autoregressive Video Generation▲14k3.00
[PDF]TraSURE: Exploiting Cross-trajectory Recurrence to Accelerate Information-seeking RL Rollouts over Shared Corpora▼5.8k2.33
[PDF]Planner-Breaker: Finding Informative Failures of Driving Planners with Reinforcement Learning▲5103.00
Withdrawn or desk rejected (643)
[PDF]Population Contracts for Scalable Language-Guided Multi-Robot Planning and ControlDesk rejected
[PDF]What Should I Try Next? Decision-Relevant Counterfactual Information Gain for Active World ModelsWithdrawn1.00
[PDF]Never Stop Rewiring: Safe Continual Reinforcement Learning on Non-Resetting GraphsWithdrawn1.33
[PDF]Recoverable Coordination Distillation for Decentralized Multi-Agent Reinforcement LearningWithdrawn1.67