EPISODE · Sep 11, 2026 · 27 MIN
[人人能懂AI前沿] 从自我进化、系统协作到负向学习:AI智能的别样路径
from AI可可AI生活
AI如何才能学会自我进化,最终成为自己的师傅?为什么解决顶级难题要靠“AI专家团”,而不是一个超级大脑?本期节目,我们将从几篇最新论文出发,探讨AI如何从别人的失败中汲取智慧,看懂“剩饭”为何难倒英雄汉,并理解“看得懂”与“会动手”之间那道巨大的鸿沟。00:00:26 那个“笨”徒弟,正在悄悄学会自己当师傅00:07:02 AI解题的秘密,不是一个大脑,而是一套系统00:11:44 AI的“偏食症”,为什么聪明的模型更讨厌“剩饭”?00:17:21 人工智能看得懂,但不会干00:21:56 如何变得更聪明?答案是,多看看笨蛋是怎么想的本期介绍的几篇论文:[LG] The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement[Shanghai Jiao Tong University]https://arxiv.org/abs/2609.11873 ---[AI] An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics[NVIDIA]https://arxiv.org/abs/2609.10712 ---[LG] Data Scarcity and Model Sparsity: Mixtures-of-Experts Overfit More to Repeated Data[Stanford University]https://arxiv.org/abs/2609.11917 ---[AI] MindTopo: Can Foundation Models Reason in Topological Space?[Northwestern University]https://arxiv.org/abs/2609.11900 ---[CL] Negative Self-Distillation: Learning to Reason by Avoiding Flaws[University of Virginia]https://arxiv.org/abs/2609.11699 在小宇宙查看该单集文稿
Embed this episode
Ready to play
[人人能懂AI前沿] 从自我进化、系统协作到负向学习:AI智能的别样路径
No transcript for this episode yet
Similar Episodes
No similar episodes found.