Reaching 95%-ile isn’t very impressive because it’s not that hard to do.…

As you may already know, Danica McKellar, the actress and UCLA mathemati…

T5 把翻译、分类、摘要、问答都改写成统一的文本到文本任务，但这个接口不是免费的。本文拆开 span corruption 的 corrupti…

A recurring discussion in Overwatch (as well as other online games) is w…

Van Vu and I have recently uploaded our joint paper, “Random matrices: t…

GPT-1 到 GPT-4 四次跃迁里，next-token prediction 这个目标函数从未换过，真正变化的是任务接口、参数规模和对齐机…

There’s a cocktail party version of the efficient markets hypothesis I f…

I recently discovered a CSS hack which automatically makes wordpress pag…

BERT 的每个设计选择都有可核对的实测代价。本文钉住 BERT 这一条路线本身：80/10/10 掩码策略真正缓解了多少 pretrain-f…

This is a response to the following question from David Albert: