Aritra Ghosh

AI Research @ Meta (MSL)

[Aritra Ghosh @_arghosh]

@LiamFedus @Max_A_Kaufmann Wouldn’t it be better to treat the action as generating a set of plausible ideas, especially when verifiable reward is very sparse or delayed? Otherwise RL mainly learns to chase what gets verified, not what was the best thinking given the evidence at the time.

[Tanay Debnath @Tanay8690]

Hi
@LatinXChem,
this is my work "Design and Application of Unnatural Base Pairs (UBPs) for Genetic Alphabet Expansion toward Semi-Synthetic Organism: A Classical MD approach" #LatinXChem22 #LXChemBio #BIO116 #MD #PolarizableForceField
@CisnerosRes

[Erik Meijer @headinthebox]

That's my team. It was a true privilege to support such an amazing group of talented individuals.

(slightly outdated public overview page https://t.co/5cW8s04GPh)