BridgeMind On Muse Wonder: I've Never Seen A Model This Benchmaxxed In My Life

Explain Like I'm Normal

Meta's new Muse Wonder model is facing heavy community criticism for 'benchmaxxing'—optimizing specifically for leaderboard scores while failing basic real-world prompts. Testers found that despite high scores, the model struggles with simple spatial reasoning like generating a lava lamp, producing only a red cylinder instead. This highlights a growing rift between synthetic performance metrics and actual model utility.