EPISODE · Jun 8, 2026 · 28 MIN
[人人能懂AI前沿] 驯服“神兽”指南:给AI纠错、开小灶与装个“省钱”的脑子
from AI可可AI生活
你有没有想过,AI不仅会犯错,犯错时还分“执迷不悟”和“一路迷茫”两种性格?我们想给AI“开小灶”教点新东西,最有效的方法竟然是发出比主信号弱一千倍的“悄悄话”。本期节目,我们将一起钻进AI的大脑,看看它是如何通过“搭便车”学坏,如何被装上一个“精打细算”的省钱脑子,以及我们该如何用几何“画圈”的方式,真正看懂它的所思所想。准备好了吗?让我们马上出发!00:00:34 AI“学坏”,竟然是因为一个“搭便车”的坏习惯?00:06:26 AI犯错,也分“执迷不悟”和“一路迷茫”?00:10:44 AI进阶的艺术,如何给它开个“小灶”?00:16:15 给AI装一个“省钱”的脑子00:22:22 AI的“脑补”和我们的“理解”,中间差了什么?本期介绍的几篇论文:[CL] The Piggyback Hypothesis of Generalization: Explaining and Mitigating Emergent Misalignment [Northeastern University & Stanford University] https://arxiv.org/abs/2606.06667 ---[CL] How Language Models Fail: Token-Level Signatures of Committed and Persistent Reasoning Failures [Stanford University] https://arxiv.org/abs/2606.06635 ---[LG] TALAN: Task-Aligned Latent Adaptation Networks for Targeted Post-Training of Large Language Models [Meta AI] https://arxiv.org/abs/2606.06902 ---[LG] Towards Tight Bounds for Streaming Attention [MIT] https://arxiv.org/abs/2606.07205 ---[LG] A Geometric View for Understanding Concept Learning and Neuron Interpretation in Sparse Autoencoders [University of Washington] https://arxiv.org/abs/2606.07007 在小宇宙查看该单集文稿
Embed this episode
Ready to play
[人人能懂AI前沿] 驯服“神兽”指南:给AI纠错、开小灶与装个“省钱”的脑子
No transcript for this episode yet
Similar Episodes
No similar episodes found.