I finally got around to making a tool to compare completions from SFT vs. RLHF trained models. This is a mini site for the RLHF book that I've wanted for a while.

buff.ly/lqDL5wa

It's always been hard to say what RLHF does to a model within a more complex post-training pipeline.

I am afraid to report, RL works.

I think 2-3 years ago, I said I will not work on two ML sub-areas. RL was one of them. I am happy to say that I am not strongly attached to my beliefs.

Latest polling for the German election has CDU/CSU leading with ~30%,
AFD second with ~ 20%,
then Greens and SPD vying for third place (SPD leading atm with ~ 16%).
FDP looks increasingly less likely to make it into parliament.

But lots of uncertainty (and look at YouGov outlier numbers
right)

Masked Autoencoders are PDE Learners

Anthony Zhou, Amir Barati Farimani

Action editor: William Redman

https://openreview.net/forum?id=rZNuiFwXVs

#autoencoders #pdes #pde