Meta explains Muse Wonder's Active Memory and Active Focus for long-running agents
AIatMeta · x · 2026-10-13
Meta AI highlights two mechanisms making Muse Wonder suited for long-running, multi-step agent work: Active Memory carries your context across conversations and surfaces it when relevant, while Active Focus keeps the model working toward your goal through blockers and alternative approaches.
On the internal FocusBench (realistic multi-step tool-use and file-processing tasks), Muse Wonder Full scores 54.2%, ahead of GPT-6 Astra (47.7%), Claude Opus 5.5 (45.8%), and Claude Fable 5.1 (36.6%).
Related event: Meta unveils Muse Wonder, its strongest model yet for personal agents(76 posts)→
More from coding & agent
- Enterprise AI's Worst Failure Mode: Confident Wrong Answers, No Errors — DavidLinthicum · 2026-10-13
- Dev builds Felix, a detective game living on an M5Stack device with Codex and receipt printer clues — OpenAIDevs · 2026-10-13
- Muse Wonder tops SWE-Marathon at 57.5%, beating Claude and GPT-6 by 11+ points — jyangballin · 2026-10-13
- 72% of ReadMe's docs traffic is agents; Restless puts API fixes inside error responses — JaynitMakwana · 2026-10-13
- Mochi: an open-source desktop companion that watches your Claude Code agents for you — brute984 · 2026-10-13
- Perplexity built Projects' filesystem as a Git forge for agents, with greps beating semantic search — perplexity_ai · 2026-10-13