[人人能懂] 给AI做脑CT、建记忆宫殿，再教它自学成才

今天，我们将深入AI的“内心世界”，看看它如何变得更聪明。我们会用最新论文中的方法，给AI做一次“脑CT”看清能力升级的代价，并教会“音盲”的它“脑补”出声音的质感。接着，我们将揭示AI如何像搭建“记忆宫殿”和使用“信息压缩机”一样，告别遗忘和臃肿。最后，我们将见证AI如何摆脱人类老师，通过预测作者思路实现“自学成才”！00:00:31 给AI做一次“脑CT”：排行榜之外，我们如何看透模型的真本事？00:05:50 AI的“记忆宫殿”：聊得再久，它怎么才能记住重点？00:11:19 给AI装上“压缩饼干”机：信息再多，也能秒懂重点00:16:27 AI学会了“脑补”声音：闭上眼睛，如何听懂全世界？00:21:44 AI界的“自学成才”：没有老师，如何炼成绝世武功？本期介绍的几篇论文：[CL] Beyond the Leaderboard: Understanding Performance Disparities in Large Language Models via Model Diffing  [HBKU]  https://arxiv.org/abs/2509.18792  ---[CL] EpiCache: Episodic KV Cache Management for Long Conversational Question Answering  [Apple]  https://arxiv.org/abs/2509.17396  ---[CL] CompLLM: Compression for Long Context Q&A  [Amazon]  https://arxiv.org/abs/2509.19228  ---[CL] AuditoryBench++: Can Language Models Understand Auditory Knowledge without Hearing?  [Pohang University of Science and Technology & HJ AILAB]  https://arxiv.org/abs/2509.17641  ---[CL] Reinforcement Learning on Pre-Training Data  [Tencent]  https://arxiv.org/abs/2509.19249  

There are no comments yet.

Please log in to write the first comment.