[人人能懂] 向AI偷师：让你变聪明的四个思维模型

00:00:32 AI也会“偏科”？高手如何跳出舒适区 00:05:55 AI进化论：从“伸手党”到“高手的秘密” 00:10:42 AI的“体检”新思路：如何看穿一个模型的“小心思”？ 00:15:17 一百万学生教会我们的事：简单，可能就是最优解 本期介绍的四篇论文：[LG] RL-PLUS: Countering Capability Boundary Collapse of LLMs in Reinforcement Learning with Hybrid-policy Optimization  [Tongyi Lab, Alibaba Group & Peking University]  https://arxiv.org/abs/2508.00222  ---[LG] MetaAgent: Toward Self-Evolving Agent via Tool Meta-Learning  [BAAI]  https://arxiv.org/abs/2508.00271  ---[LG] Watch the Weights: Unsupervised monitoring and control of fine-tuned LLMs  [Carnegie Mellon University (CMU)]  https://arxiv.org/abs/2508.00161  ---[LG] Learning to Optimize Feedback for One Million Students: Insights from Multi-Armed and Contextual Bandits in Large-Scale Online Tutoring  [Carnegie Mellon University (CMU) & CK-12 Foundation]  https://arxiv.org/abs/2508.00270  在小宇宙查看该单集文稿