评分 6 · 来源：cs.CL updates on arXiv.org · 发布于 2026-04-17
评分依据：Scales MATS for code generation via RL, addresses scalability issue in multi-agent code generation
arXiv:2604.14564v1 Announce Type: cross Abstract: Reinforcement learning (RL) paradigms have demonstrated strong performance on reasoning-intensive tasks such as code generation. However, limited trajectory diversity often leads to diminishing returns, which constrains the achievable performance ceiling. Search-enhanced RL alleviates this issue by introducing structured exploration, which remains constrained by the single-agent policy priors.