Xiang Yue (@xiangyue96)

There are competing views on whether RL can genuinely improve base model's performance (e.g., pass@128). The answer is both yes and no, largely depending on the interplay between pre-training, mid-training, and RL. We trained a few hundreds of GPT-2 scale LMs on synthetic GSM-like reasoning data fro...

Demystifying Long CoT Reasoning in LLMs <a target="_blank" href="https://arxiv.org/pdf/2502.03373" color="blue">arxiv.org/pdf/2502.03373</a> Reasoning models like R1 / O1 / O3 have gained massive attention, but their training dynamics remain a mystery. We're taking a first deep dive into understandi...