Hyun Lee

Flat-RLVR: Sharpness-Aware Reinforcement Learning with Verifiable Rewards for Reasoning Capacity in LLMs

Reinforcement learning with verifiable rewards (RLVR), particularly through group-relative policy optimization (GRPO), has emerged as a central paradigm for the post-training of large language models (LLMs). While RLVR is highly effective at enhancing the reasoning capabilities of LLMs, it commonly induces \textit{dist...