Meta Distills Flagship Muse Wonder Into Daily Assistants, Cutting Forgetting 2.4x

Meta uses its flagship Muse Wonder models as training sources to retrain its everyday Muse and Meta AI assistants, cutting forgetting by 2.4x, with the updated models already rolling out to users.

2026-11-09 ~ 2026-11-09 · 2 related posts

• Meta retrains its everyday models on flagship Muse Wonder: 2.4x less forgetting, +67% task completion — xiaohu · 2026-11-09
• Deep dive: how Meta's Muse Wonder retraining cut assistant forgetting by 2.4x — xiaohu · 2026-11-09