AI RESEARCH
Beyond pass@k: Redundancy-Aware RLVR for Multi-Sample Code Generation
arXiv CS.CL
•
ArXi:2605.28022v1 Announce Type: new LLMs for code generation are commonly evaluated in repeated-sampling settings using Pass, where multiple candidate programs are executed against unit tests under a finite sampling budget. While recent verifier-based reinforcement learning (RLVR) methods improve executable correctness, how these objectives affect redundancy among sampled programs remains poorly understood. In this work, we study implementation-level redundancy in code generation using JPlag, a plagiarism-detection system for code.