Can reinforcement learning for LLMs scale beyond math and coding tasks? Probably

https://arxiv.org/abs/2503.23829

#HackerNews #reinforcementlearning #LLMs #scaling #math #codingtasks #AIresearch