Seonglae ChoAIT
AIT Notion
Singular Learning Theory extends AIT to provide new prediction error bounds that explain everything from Bayesian learning in neural networks to OOD generalization
From SLT to AIT: NN generalisation out-of-distribution — LessWrong
TL;DR: This post derives an upper bound on the prediction error of Bayesian learning on neural networks. Unlike the bound from vanilla Singular Learn…
https://www.lesswrong.com/posts/2MX2bXreTtntB85Zy/from-slt-to-ait-nn-generalisation-out-of-distribution
