Dave Graham

davegraham.bsky.social

01 April 2026 17:22

@mlcommons.org maintains a few repositories of people’s speech for AI training (which was done consensually and is made open). I’m sure there’s places that would love to maintain this archive if provided.

Scott Wasson

scottwasson.bsky.social

01 April 2026 17:22

MLPerf Inference 6.0 results are out! Lots of new goodness here, including a bunch of new benchmark tests (and models) and some very large clusters reaching new peak performance numbers.

Elena Simperl

elenasimperl.bsky.social

20 February 2026 00:45

So proud to co‑chair this work. It has been almost two years in the making and I'm grateful to the many brilliant colleagues who contributed their ideas, feedback, and energy.

eleutherai.bsky.social

20 February 2026 00:45

Announcing our latest paper: CommonLID In collaboration with @commoncrawl.bsky.social @mlcommons.org @jhu.edu we built a LID benchmark on actual Common Crawl text covering 109 languages. Existing evaluations overestimate how well LangID works on web data. arxiv.org/abs/2601.18026

Dave Graham

davegraham.bsky.social

04 February 2026 18:29

We’re doing testing on jailbreaking most of the modern Agentic interfaces today at @mlcommons.org. You can take a look at some of our artifacts here: mlcommons.org/ailuminate/j... Suffice it to say, these types of exploits are real (as you’ve noted), very dangerous (again, noted) & persistent