Latest

Creating artificial doubt significantly improves AI math accuracy

Step-by-step reasoning can fix madman logic in vision AI

All LLMs use tokenization. Are we doing it totally wrong?

Slashing model size by 85% while redefining how we build adaptable, efficient LLMs

Can image models understand what we’re asking for?

High-quality graphics vs high-quality understanding — which one matters more?

Teaching AI to tell visually consistent stories

Bye tokens, hello patches

Meta announces a better way to scale LLMs

Input prompt, output a playable world

Google built a world model that actually works

ChatGPT naturally colludes to raise prices

Making AI See Better

Chain-of-thought is so hot right now. You shouldn't always use it.

Get ready to lose to Transformers on Lichess

They can hit 2895 Elo … without memorizing patterns

Long Context Compression with Activation Beacon

Differential Transformers

LLMs will lie forever

Hallucinations are never going away. How can we reduce them?