Mathy AI Substack

Mathy AI Substack

Share this post

Mathy AI Substack
Mathy AI Substack
An unusual look of the KL-Divergence Term in Deep Seek R1 Training Objective?

An unusual look of the KL-Divergence Term in…

Mike Erlihson, Mathy AI
Jan 28
5

Share this post

Mathy AI Substack
Mathy AI Substack
An unusual look of the KL-Divergence Term in Deep Seek R1 Training Objective?

DeepSeek Blog Series: based on http://joschu.net/blog/kl-approx.html

Read →
Comments
User's avatar
© 2025 Mike E.
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture

Share