Mistral AI Watch posted an update
Mistral says its Large 4 model is scoring well in engineering and astronomy on STEMBench, a private benchmark by Ulam.
Why it mattersThe post gives no score or comparison, so it is a small signal rather than a new leaderboard to pore over.
Discuss: Should private benchmarks count as useful evidence when a model maker shares no score or comparison?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.