Meta's Muse Wonder Full scores just 1% on MazeBench, near the bottom of the board
patience_cave · x · 2026-10-14
On the long-horizon MazeBench benchmark, Meta and partners' Muse Wonder Full scored just 1%, with Muse Wonder at 0%.
- The leaderboard screenshot shows gpt-6-astra (Codex max) leading at 14%, followed by gpt-6.1-sol at 9% and claude-opus-5-5 at 6%
- Muse Wonder Full sits at the 1% mark alongside several Claude models
- The poster framed it sarcastically as an "impressive entry by Meta and friends"
More from Models
- Opus 5.5 still leads a month later, but open-weight decision models are closing in — airesearch12 · 2026-10-16
- Google restricts Gemini Pro access for Google Plus subscribers, matching free tier — EmergencyLess8861 · 2026-10-16
- Meta releases Muse Spark Open, trained with vLLM-powered RL inference from day one — colinmcnamara · 2026-10-16
- Meta open-sources Muse Spark: runs on two Mac Studios, license bans passing output off as human-written — mawcopolow · 2026-10-16
- Golden Gate Claude explainer resurfaces: interpretability premise still holds — HarperSCarroll · 2026-10-16
- Hugging Face livestreams release of a new open-weight model — huggingface · 2026-10-16