首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Lai Wei · Liangbo He · jun lan · Lingzhong Dong · Yutong Cai · Siyuan Li · Huijia Zhu · Weiqiang Wang · Linghe Kong · Yue Wang · Zhuosheng Zhang · Weiran Huang
Chen Wang · Lai Wei · Yanzhi Zhang · Chenyang Shao · Zedong Dan · Weiran Huang · Ge Lan · Yue Wang
Recent advances in reinforcement learning (RL) have improved the reasoning capabilities of large language models (LLMs) and vision-language models (VLMs). However, the widely used Group Relative Policy Optimization (GRPO) consistently suffers from entropy collapse, causing the policy to converge...