Re:Gato – A Generalist Agent
by abel_ on Tuesday May 17, 2022 @08:23PM
(#31417808)
There's a neat argument against these models doing interpolation: the manifold of the data is so sparse that it's infinitesimally unlikely for a good predictor to be doing interpolation between existing points on the manifold.
https://arxiv.org/abs/2110.09485