Muse Wonder Full beats its scaling trend on out-of-distribution tasks, Meta says
AIatMeta · x · 2026-10-13
Meta AI compared test-time compute scaling on tasks seen during post-training versus unseen tasks for the Muse Wonder models.
Both improve similarly in-distribution, but Muse Wonder Full's out-of-distribution score exceeds what the Muse Spark 1.3 trend predicts — an upward shift Meta attributes to a stronger base model that carries post-training gains over more effectively to unanticipated tasks.
Related event: Meta unveils Muse Wonder, its strongest model yet for personal agents(76 posts)→
More from Models
- Meta's rumored Muse Wonder beats Claude Opus 5.5 on several agent and coding benchmarks — Dr_Singularity · 2026-10-13
- Wonder Full is a good model, but Meta still trails OpenAI by ~3-6 months — scaling01 · 2026-10-13
- Muse team member celebrates coding work hitting SOTA 'again and again' — jyangballin · 2026-10-13
- Muse Wonder tops SWE-Marathon at 57.5%, beating Claude and GPT-6 by 11+ points — jyangballin · 2026-10-13
- 60.6% of LLM theorem rewrites silently drop hidden hypotheses, study finds — MFarajtabar · 2026-10-13
- Meta's Wonder Full Debuts #5 on Vals Index, Outranking Every OpenAI Model — airesearch12 · 2026-10-13