Digest | Garden
All Bookmarks 3319

@Saket Gokhale

Saket Gokhale: Finally!

Feb 15

@Aran Komatsuzaki

Apple presents: Distillation Scaling Laws Presents a distillation scaling law that estimates...

Feb 15

@McKay Wrigley

This is my o1-pro workflow guide for building apps from a starting template.

Feb 15

@Kevin Henrikson

Enter Goose: A free, open-source framework that simplifies the process of building AI agents

Feb 13

@Balaji S. Srinivasan

Carl is a Bitcoin core developer with years of experiencing building VPNs that can punch through...

Feb 12

@Noam Brown

This meme summarizes the paper nicely

Feb 12

@Mahesh Sathiamoorthy

Some people asked me about a resource for learning about Transformers.

Feb 10

@David

2.3k stars on this github repo in 3 days... wat

Feb 10

@John Chen

GauntletAI Week 2 Day 2: Seeing what Spiderman and Ironman think about Trump AKA Making an...

Feb 8

@Oriol Vinyals

Distillation has been on the news (!) due to @deepseek_ai. The paper was actually rejected from...

Feb 8

@MMitchell

MMitchell: New piece out!

Feb 8

@Oskar Stålberg

So whats your best algorithm for generating aperiodic infinite deterministic irregular relaxed...

Feb 8

@Sahil Lavingia

No longer hiring junior or even mid-level software engineers.

Feb 7

@Franck SN

One underrated feature is that by adopting @LiteLLM as a 'backend,' DSPy now makes it easy to...

Feb 2

@Franck SN

Just created a new LLM leaderboard using DSPy, Chess, and @nebiusai. The results gave me food for...

Feb 2

@Franck SN

@karpathy This is exactly what the DeepSeek paper shows. I expect a lot of gains from RL in the...

Feb 2

@Omead Pooladzandi

We've been cooking.

Jan 31

@AmishMagic

Like everyone, I've been a bit distracted exploring @deepseek_ai R1 and experimenting with it...

Jan 31

@Avichal Garg (Electric Capital) ⚡

1/ This screenshot is floating around and I know the back story.

Jan 31

@Om Alve

So I pre-trained an LLM from scratch this week, yes, pretrained not finetuned. This is how it went...

Jan 31

@Wordgrammer

Okay. Thanks for the nerd snipe guys. I spent the day learning exactly how DeepSeek trained at 1/30...

Jan 29

@Noahdgoodman

Congrats to OAI on producing a reasoning model! Their opaque tweets demonstrate that they’ve...

Jan 29

@Abhinav Upadhyay

End-to-end ML School was a wonderful learning resource. Probably one of the earliest tutorials on...

Jan 29

@N8 Programs

reading a deepseek paper and stumbled upon a very beautiful formula where they *unify SFT and MOST...

Jan 29
auto_stories

Select a bookmark to read

Click any item from the list