For the third installation of Spooktober Optimizers, health tech companies are getting a bit... vampiric. I shipped off 24 vials of blood this year for testing — 8 in the last six weeks for work. What did I get out of it? A vague promise of maybe living longer. 🎁: tinyurl.com/m3mvt9px

Show Post

A huge loss for science and health reporting. Thank you for all your critical work @helenbranswell.bsky.social!

Show Post

Aww, thank you Sophie!

Show Post

In a new essay, David Hall @dlwh.bsky.social argues that pretraining consumes documents, while posttraining consumes tasks. Open language models could run on Wikipedia and ArXiv. Posttraining has no equivalent commons. David reflects on the choice we get to make: openathena.ai/blog/there-i...

Show Post

Marin's 535B-parameter hero run was originally planned as a 360B model. A new blog from Rafal Wojdyla, MTS, explains how a change in sharding strategy, Expert Parallelism, made room for a ~50% larger model at the same ~23B active parameters, and what it costs. 🔗 openathena.ai/blog/expert-...

Show Post