@alexandr_wang 1/ Muse Spark 1.1 just dropped into #3 on the Debate Benchmark. Behind only Claude Fable 5 and Opus 4.7, ahead of GPT-5.6 Sol. Not bad for Meta’s “we pivoted from open weights” era. 2/ What makes this interesting isn’t just the ranking. It’s the pattern: Meta is no longer
Full content is available at the original source.
x.com