TimDarcet

AI at Meta Verified account @AIatMeta · Aug 10

Introducing Muse Glimmer, an open-weight 30B-parameter model optimized for local, always-on agent workflows.

Muse Glimmer delivers strong performance on key agentic use cases and benchmarks compared with leading models in its size category, and is designed to run entirely on consumer hardware like a Mac or PCs with performant GPUs.

In keeping with our long tradition of sharing fundamental AI research, we’re releasing model weights under a permissive Apache 2.0 license.

🧵👇

View quoted post on X

nightwing Verified account @yaboilyrical · Oct 28

L6N5401-0.75 seems to be the 'DJ neuron' - it puts everything into the context of a DJ set

View quoted post on X

Yann LeCun Verified account @ylecun · Nov 5, 2016

Looks like we have over 500 submissions to ICLR this year.

openreview.net/group?id=ICLR.…

View quoted post on X

AI at Meta Verified account @AIatMeta · Oct 19

Today we published a technical report on Muse Glimmer, our 30B open-weight model built for agentic tasks on consumer hardware.

It runs at practical speeds on a Mac or PC with a single consumer GPU at 4-bit precision, and leads comparably-sized open models on tool use and multi-step planning.

Read about the model architecture, distillation recipe, quantization and more: go.meta.me/4Bcvxol

View quoted post on X

Alan Baade Verified account @BaadeAlan · Oct 19

1/ The tokenizer is a new scaling axis!

We introduce an approach for diffusable representation learning that's principled, scalable, and works when DINO/LeJEPA collapse.

We get joint tokenizer+diffusion scaling laws and it's compute-optimal vs raw diffusion

w/ @drfeifei @GordonWetzstein @jcjohnss @ericryanchan @vincesitzmann @KyleSargentAI

Thread, Paper, and Blogpost below

View quoted post on X