Context compression finally works in production: new research cuts LLM input 16x without the accuracy hit

LCLMs compress LLM context before decode — 8.8x faster at 16x compression, beating every KV cache method tested. Open-sourced by NYU and Columbia.

Google's DiffusionGemma generates 256 tokens in parallel and self-corrects as it goes

Google's open-source diffusion language model generates 256 tokens in parallel and self-corrects, hitting 4x speed on one GPU ...

Cleve Moler, Who Unlocked the Power of Computing for Millions, Dies at 86

He built interfaces that allowed engineers, scientists and everyday people to solve difficult problems without having to ...