tracked author

Yuxian Gu

LLM-in-Sandbox coauthor from Tsinghua University. Homepage identifies him as a final-year PhD candidate in the Conversational AI Group at Tsinghua University, with work on LLM efficiency across pre-training, adaptation and inference; he has also worked with Microsoft Research Asia and MIT HAN Lab.

1 archived notes X: not-found Homepage

Related Notes

按论文归档时间排序,展示该作者在本站已经出现的材料。

归档

Computer Environments Elicit General Agentic Intelligence in LLMs

LLM in Sandbox 只给模型 shell、文件编辑和完成信号,使部分模型—任务组合获得最高 15.5 个百分点增益,并让 Qwen3 4B 经文件型通用任务强化学习后把平均交互轮次从 23.7 降到 7.0,这些数值来自允许联网的特定环境配置。

待审阅 2601.16206-computer-environments-agentic-intelligence Tool UseAgent RLAgent Workflow