Sign inSign in

Nati Srebro is a Professor at UC Berkeley working on statistical learning theory and optimization for machine learning.

More

Nati Srebro is a Professor at UC Berkeley working on statistical learning theory and optimization for machine learning. In a research-level Simons Institute talk, Srebro shows how chain-of-thought decouples transformer runtime and memory from model size.

Playlists

1

About

Nati Srebro is a Professor at UC Berkeley working on statistical learning theory and optimization for machine learning. In a research-level Simons Institute talk, Srebro shows how chain-of-thought decouples transformer runtime and memory from model size.

Connections

10

Their places, courses, books and lectures.