智能AI None 使用级联线性特征检测和控制阿谀奉承 arXiv:2606.26155v1 Announce Type: new Abstract: Interpreting and controlling model behaviors through activation steering and that behavior 2026-06-26 Maty Bohacek, Rishub Jain, Nicholas Dufour, Thomas Leung, Chris Bregler, Roma Patel
智能AI None GROW：将 GRPO 与开放世界 VLM 代理的状态动作建模结合起来 arXiv:2605.20246v1 Announce Type: new Abstract: Recently, vision-language model (VLM) agents have shown promising progre the samples VLM 2026-05-21 Xiongbin Wu, Zhihao Luo, Shanzhe Lei, Lechao Zhang, Xuhong Wang, Jie Yang, Zhonglong Zheng, Yuanjie Zheng, Xin Tan, Wei Liu