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
125 results for “keyword:"vr"” in ICLR 2027. 23 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]RLVR as Prompt Augmentation reinforces both Correct and Incorrect Reasoning to the Correct Answer▼4.3k2.00
[PDF]What Stochastic Information Matters for Learning? Observable-Relative Equivalence and Control▲3.0k2.33
[PDF]When Clipping Becomes a Learning Signal: A Permutation Baseline for Policy Optimization▲3.0k2.33
[PDF]When Do RTL Verifiers Lie? Measuring Proxy-Reward Overoptimization in LLM Hardware Design▼4.5k2.67
[PDF]PROSE: Enhancing Sample Efficiency in Code Generation via Plan-Aware Rollout Reorganization▼5.8k2.33
[PDF]Audit Certainty Is a Signed Lever: Why a Verifier Calibrated on a Proxy Can Backfire in Deployment▼5.8k2.33
[PDF]Count the constraints, not the tests: verifier diversity governs reinforcement learning from execution feedback▼1.3k1.75
[PDF]How Fast Should a Model Commit to Supervision? Training Reasoning Models on the Tsallis Loss Continuum▲3952.67
[PDF]Exploration Prior: Semantically Guided Exploration for Reinforcement Learning of Language Model Reasoning▼4.2k2.20
[PDF]CKA's Blind Spot: Why Representation Analysis Misses Policy Change in RLVR, and How to Fix It▼1.3k1.75
[PDF]Correct Reward, Costly Rule: Gradient Coherence Decides What Group-Relative RL Locks In▼1.3k1.75
[PDF]Same Mastery, Opposite Directions: Signed Learning Progress for Prompt Revisiting in RLVR▼1.2k1.67
[PDF]Diagnosing Recoverability: Suppression-Like vs. Overwrite-Like Capability Loss in LLM Post-Training▲3952.67