-
「Open question: boundedness of the trilinear Hilb」：数学直觉如何落到可检验的步骤
This is a well-known problem in multilinear harmonic analysis; it is fas…
-
测试投入很少时，应该怎么测软件
I’ve been reading a lot about software testing, lately. Coming from a ha…
-
跟踪公开讨论两年：研究信号怎么从噪音里捞出来
In early 2022 I realized that being a good researcher isn’t just about d…
-
「Distinguished Lecture Series I: Shing-Tung Yau,」：数学直觉如何落到可检验的步骤
The final Distinguished Lecture Series for this academic year at UCLA wa…
-
2026AI文本水印实测揭秘：普通人完全无法分辨水印文本差异
2026年主流大模型落地的AI文本水印（以SynthID-Text为核心），并不会降低文案输出质量、不会泄露用户隐私，也无法被普通人肉眼识别。4…
-
终端延迟从哪来
There’s a great MSR demo from 2012 that shows the effect of latency on t…
-
关于大语言模型的六个工程直觉：下一词预测为什么能学会那么多
An open question these days is why large language models work so well. I…
-
读懂「Distinguished Lecture Series II: Shing-Tung Yau,」：先把定义、反例和适用范围钉死
On Thursday, Yau continued his lecture series on geometric structures, f…
-
市场、歧视和「降低门槛」
Public discussions of discrimination in tech often result in someone cla…
-
语言模型评测怎样才算成功：能指导训练，而不是只刷榜
Everybody uses evaluation benchmarks (“evals”), but I think they deserve…