Reinforcement Learning with Verifiable yet Noisy Rewards under Imperfect Verifiers
cs.LG, cs.AI
Submitted: 2025-10-01
Updated: 2026-10-05
Code: https://github.com/caixq1996/noisy-RLVR
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Evaluating Large Language Models Trained on Code
- Training Verifiers to Solve Math Word Problems
- A Survey on LLM-as-a-Judge
- From Accuracy to Robustness: A Study of Rule- and Model-based Verifiers in Mathematical Reasoning
- On the admissibility of Horvitz-Thompson estimator for estimating causal effects under network interference
- VerifyBench: A Systematic Benchmark for Evaluating Reasoning Verifiers Across Domains
- Reinforcement Learning with Verifiable Rewards: GRPO's Effective Loss, Dynamics, and Success Amplification
- Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges
- Reinforcement Learning with Verifiable Rewards Implicitly Incentivizes Correct Reasoning in Base LLMs
- TinyV: Reducing False Negatives in Verification Improves RL for LLM Reasoning
- One Token to Fool LLM-as-a-Judge
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks