Bryan Bischof fka Dr. Donut (@bebischof) 's Twitter Profile
Bryan Bischof fka Dr. Donut

@bebischof

Leading AI @theoryvc; Teach ML @rutgersu. Prev: Led AI @_hex_tech; Led Data @weights_biases, ML @stitchfix, Led Data @bluebottleroast

ID: 223231567

linkhttps://bio.site/Docdonut calendar_today05-12-2010 19:56:01

21,21K Tweet

4,4K Followers

2,2K Following

Jason Ganz (@jasnonaz) 's Twitter Profile Photo

Bryan Bischof fka Dr. Donut - Most diseases are solved - We have access to untapped knowledge on demand When you arrive to give a presentation at a conference or meetup the slides won't display reliably and you'll have to fiddle with 15 different settings to get it to work

Bryan Bischof fka Dr. Donut (@bebischof) 's Twitter Profile Photo

I, for one, am completely white-pilled on this. At Hex last year, we had a sub-group on the AI eng team called 'context construction' and that team's operating principles were: - lots of experiments - operating over clear evals - start explicit, rollup functionality over time

Jason Ganz (@jasnonaz) 's Twitter Profile Photo

Anyone know any great visual artists / designers who would be interested in a paid project to help create a unique physical artifact?

Bryan Bischof fka Dr. Donut (@bebischof) 's Twitter Profile Photo

If you read this and think “that’s neat, I bet this kind of work is fun”, we should chat. My Eng team at Theory VC is doing projects just like this, across the entire venture operations stack: deal flow, research, support, etc.

Bryan Bischof fka Dr. Donut (@bebischof) 's Twitter Profile Photo

Such an amazing visual demonstration! I really want to rewrite some of the findings from the MindAgents work in the language of this paper!

Hamel Husain (@hamelhusain) 's Twitter Profile Photo

Overview of the series 1. We’ve been measuring wrong. Nandan Thakur showed that traditional IR metrics optimize for finding the #1 result. RAG needs different goals: coverage (getting all the facts), diversity (corroborating facts), and relevance. Models that ace BEIR benchmarks

Overview of the series

1. We’ve been measuring wrong. <a href="/beirmug/">Nandan Thakur</a> showed that traditional IR metrics optimize for finding the #1 result. RAG needs different goals: coverage (getting all the facts), diversity (corroborating facts), and relevance. Models that ace BEIR benchmarks
Bryan Bischof fka Dr. Donut (@bebischof) 's Twitter Profile Photo

"Oh your office is in the Ferry building, that's so amazing, what's it like" "Non-stop busking. Meetings interrupted by bagpipers outside. Front row seats to the next Santana in the streets."