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
211 results for “keyword:"verifiable rewards"” in ICLR 2027. 31 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]The Protocol Is the Agent: Decomposing RL, SFT, and Scaffold Effects in Tool-Use Agents from 0.8B to 27B▼1.2k1.67
[PDF]RACS: Response Activation Consistency-guided Data Selection for Reinforcement Learning with Verifiable Rewards▼3.7k2.25
[PDF]Correct Reward, Costly Rule: Gradient Coherence Decides What Group-Relative RL Locks In▼1.3k1.75
[PDF]BlendPO: Controlling Guidance Strength in Reinforcement Learning without Changing Guidance Content▲9.2k2.67
[PDF]Spectral-Driven Adaptive-Rank Optimizer for Large Language Model Reinforcement Learning with Verifiable Rewards▲3.0k2.33
[PDF]UltraLogic: Enhancing LLM Reasoning through Large-Scale Data Synthesis and Bipolar Float Reward▼1.2k1.67
[PDF]Reinforcement Learning In Partially Verifiable Environments for Colon Cancer Treatment Planning▼4.3k2.00
[PDF]WHICH VERIFIER ERRORS REACH THE GRADIENT? A MOMENT DECOMPOSITION FOR AUDITING GROUP- RELATIVE REINFORCEMENT LEARNING▼5.8k2.33
[PDF]CARVE: Coordinating Imitation and Exploration through Reward Variance for LLM Reasoning Post-Training▼1361.33
[PDF]Measure First, Then Allocate: Calibrated Rollout Budgets for MCP Agents in Reinforcement Learning▼4.3k2.00
Withdrawn or desk rejected (31)
[PDF]Policy-Verified Self-Training: Learning Agent Compliance from Verifiable Policy RewardsWithdrawn1.67
[PDF]Adapt Only When It Pays: Selective, Verifiable Test-Time Self-Training for LLM ReasoningWithdrawn1.00
[PDF]Better Samples, Wrong Odds: When Prediction Rewards Distort World-Model ProbabilitiesWithdrawn1.67
[PDF]WHAT SHOULD A SELF-DISTILLATION TEACHER SEE? SOLVER FEEDBACK FOR LOGICAL REASONINGWithdrawn1.33
[PDF]Gold-LP: Data Selection for Vision-Language Reinforcement Learning with Verifiable RewardsWithdrawn1.67
[PDF]Read Together, Write Apart: Decoupling Rollout Access from Update Ownership in Multi-Role LLM TrainingDesk rejected2.00