EPISODE · May 29, 2026 · 31 MIN
[人人能懂AI前沿] 信号、记忆与野性:解锁AI黑箱的三把钥匙
from AI可可AI生活
你有没有想过,当AI学会了耍心机,我们怎么才能看穿它的伪装?最新论文带来了一把钥匙:给AI做个“脑CT”,直接定位它脑中的“坏心思”和“好创意”。本期节目,我们将一起探索如何用“蜜罐”测试AI的忠诚度,如何拯救一个创意枯竭的“好学生”AI,并从记忆的极限和“心有灵犀”的秘密中,窥见智能的本质。准备好了吗?让我们即刻出发!00:00:33 如果AI学会了“耍心机”,我们怎么才能发现?00:07:28 我们能给AI做个“脑CT”吗?00:12:23 如何拯救一个“无聊”的好学生?00:16:34 你的记忆力,正在如何塑造你的学习天花板?00:23:42 AI心有灵犀的秘密,藏在一根“直线”里本期介绍的几篇论文:[LG] Realistic honeypot evaluations for scheming propensity [Google DeepMind] https://arxiv.org/abs/2605.29729 ---[AI] Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet [Anthropic] https://arxiv.org/abs/2605.29358 ---[CL] Recovering Diversity Without Losing Alignment: A DPO Recipe for Post-Trained LLMs [University of Maryland] https://arxiv.org/abs/2605.30021 ---[CL] On Language Generation in the Limit with Bounded Memory [Cornell University & Stanford University] https://arxiv.org/abs/2605.30324 ---[LG] Representation Alignment Rests on Linear Structure [MIT] https://arxiv.org/abs/2605.28870 在小宇宙查看该单集文稿
Embed this episode
Ready to play
[人人能懂AI前沿] 信号、记忆与野性:解锁AI黑箱的三把钥匙
No transcript for this episode yet
Similar Episodes
No similar episodes found.