[人人能懂AI前沿] AI的“人性”弱点：当它学会偷懒、后悔与走捷径

你有没有想过，AI也会“偷懒和稀泥”，甚至在“不后悔”这件事上比我们做得更好？这一期，我们将一起揭开AI的“隐秘角落”，看看最新论文是如何让AI从一个只会算“相似度”的感觉派，变成一个懂得回溯证据链的逻辑派，并揪出它背后那个爱走捷径的“品味导师”的。

00:00:23 AI的学习悖论，从拼图到填词游戏

00:05:09 AI的“相似度陷阱”，为什么它总搞错“和”与“不”？

00:11:07 如何让AI学会“不后悔”？

00:17:01 AI对话，如何揪出每一句话的“祖宗”？

00:22:30 AI的“潜规则”，它在偷偷学什么？

本期介绍的几篇论文：

[CL] The JEPA Paradox in Language: The Geometry of Linguistic Alternatives

[VinUniversity & Mohamed bin Zayed University of Artificial Intelligence]

arxiv.org

---

[CV] Similarity Is Not Logic: Factored Inference for Dual-Encoder Vision-Language Models

[CMU]

arxiv.org

---

[LG] Training with (Swap) Regret Loss in a Single-Layer Self-Attention Model: A Case Study on the Probability Simplex

[MIT]

arxiv.org

---

[CL] Tokengeist: Multi-Turn Attribution Tracing in Agentic Conversations

[Microsoft Research & University of Toronto]

arxiv.org

---

[LG] What do Reward Models Memorize?

[University of Amsterdam & Google DeepMind]

arxiv.org