
Meta says mathematicians using Muse Spark answered five open research questions
Six papers span probability, differential equations and group theory; Meta says humans directed the work and verified the results.

The mixture-of-experts model activates 3.46B parameters per token and supports up to 1M tokens of context, per its Hugging Face card.

Six papers span probability, differential equations and group theory; Meta says humans directed the work and verified the results.

Cognition reports up to 4.8x total token throughput versus GB200 NVL72 on its SWE-2 inference workload; figures are company-measured.
The White House fact sheet covers official communications only; existing regulations, contracts and statutory definitions are unchanged.
The hand has 13 degrees of freedom from identical actuators and dense tactile sensors; the company dropped the pinky to cut cost and failure points.

Google's post lists a 1M-token output limit and $2/$10 per million token pricing; broad availability has no date yet.

The streaming transcription model covers 60 languages and is in public preview on Azure; Microsoft says it ranks first on Artificial Analysis.
Each dot gets its own cloud computer and 4,000+ app connectors; rollout starts with Pro and Business Premium users, excluding the EEA, Switzerland and UK.
The North Dakota campus is fully leased and contracted for 400 MW at full buildout, according to the company.
The hands have 13 degrees of freedom and drop the little finger to cut cost and failure points; they are not yet deployed across the Atlas fleet.

Matthew Schwartz describes the toolkit in an Anthropic guest post; the work is funded by Anthropic and the manuscripts are preprints, not yet peer reviewed.
The Cambridge Programme on AI Science & Policy report calls for auditing AI R&D automation and emergency preparation; it is an argument, not a measurement.