00:01:32 想让AI变聪明？关键不是喂饱它，而是教它如何“犯错” 

00:07:12 你的“电量焦虑”，AI来想办法了？

00:11:21 给AI装上一个“选择性失忆”的大脑

00:15:41 AI“现学现卖”的秘密，可能被揭开了   

00:21:06 AI也需要“自知之明”：我们如何教会机器谦虚？

本期介绍的几篇论文：

[LG] From Reasoning to Super-Intelligence: A Search-Theoretic Perspective  

S Shalev-Shwartz, A Shashua  

https://arxiv.org/abs/2507.15865 

---

[LG] Expert-Guided LLM Reasoning for Battery Discovery: From AI-Driven Hypothesis to Synthesis and Characterization  

S Liu, H Xu, Y Ai, H Li...  

Université de Montréal & University of Oxford  

https://arxiv.org/abs/2507.16110 

---

[CL] Beyond Context Limits: Subconscious Threads for Long-Horizon Reasoning  

H Luo, N Morgan, T Li, D Zhao...  

MIT CSAIL  

https://arxiv.org/abs/2507.16784 

---

[CL] Learning without training: The implicit dynamics of in-context learning  

B Dherin, M Munn, H Mazzawi, M Wunder...  

Google Research  

https://arxiv.org/abs/2507.16003 

---

[LG] Beyond Binary Rewards: Training LMs to Reason About Their Uncertainty  

M Damani, I Puri, S Slocum, I Shenfeld...  

MIT  

https://arxiv.org/abs/2507.16806