Vibeleaderboard
← All Intel
Intel / article

Muse Wonder

Source
research.meta.ai
Author
meetpateltech
Date

Context

Meta describes Muse Wonder, its new AI model family, as combining memory across conversations with the ability to sustain work on longer tasks. Its Active Memory feature carries about the user across conversations and retrieves the details relevant to the task at hand. Active Focus describes trying alternative approaches, working through blockers and adapting to new information, while asking for clarification and checking in before important actions. These are the company's descriptions of the capabilities. Meta's announcement makes Muse Wonder available through its model API and Muse Code, with Muse Wonder Full available to Muse Code subscribers.

Key takeaways · AI-distilled
  • Access is split: Muse Wonder is on the Meta Model API and Muse Code, while Muse Wonder Full is limited to Muse Code subscribers; Meta says Muse Wonder will also be used to develop specialized models for Muse.
  • Meta defines Active Memory as carrying important context about a user across conversations and retrieving details relevant to the current task, and calls reliable memory one of the most critical steps toward useful personal agents.
  • Active Focus covers trying alternative approaches, working through blockers and checking in before important actions; Meta says it measures this with FocusBench and long-running software work with SWE-Marathon.
  • Meta claims Muse Wonder Full does better on tasks not targeted in post-training than the Muse Spark 1.3 trend predicts, and attributes this to a larger, stronger base model whose post-training gains transfer better.
  • Meta says the models carry its strongest safeguards yet against CBRN and cyber misuse while keeping cyber capabilities for defensive use; a full report with additional safety evaluations is promised soon.
Terms in this piece · Glossary
  • context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
Read the source research.meta.ai
More from meetpateltech
Recommended reads
Comments

Checking sign-in…

Loading comments…