Sublime
An inspiration engine for ideas
Arnav Sheth
@arnavsheth
Some convex combination of financial economist, mathematician, and statistician. Recovering academic, wantrepreneur, and non-doer.
test
jared liu (孤独的二向箔) • 1 card
default
Koushik Sai • 2 cards
compression
oliver • 1 card
agency benchmark
Apolline Daimaru • 20 cards
genie3
arroh eth • 11 cards
humantech
DavidGsda • 70 cards
Leith Benkhedda
@leith
Introducing Husky Hold’em Bench, the first OS pokerbots eval!
https://t.co/glqwTyNf1t
We’ve seen lots of great work in OS recently on evaling LLMs on strategic gameplay.
Here’s another one with a twist: the model doesn’t get to choose its actions directly but... See more