1
DecompRL: Solving Harder Problems by Learning Modular Code Generation
借助强化学习让模型学会拆解复杂问题并生成模块化代码，冲击更难编程挑战的前沿研究。
arXiv:2607.02390v1 Announce Type: new Abstract: How can Large Language Models (LLMs) solve problems they currently cannot? Repeated sampling scales te…