Meta trains new models on Wonder to bring memory features to all users
Meta has deployed specialized model versions trained with Muse Wonder, granting Active Memory and Active Focus to all users while reserving the base model for complex tasks.
Key points
- Meta trained new versions of existing models using its frontier Muse Wonder system.
- Active Memory and Active Focus features are now available to all Muse and Meta AI users.
- The underlying Muse Wonder model remains reserved for harder subscriber tasks.
According to a report by The New Stack, published via Techmeme, these new models bring Active Memory and Active Focus capabilities to all users of Muse and Meta AI.reported
Evidence: reported
Meta introduces specialized versions of its existing models, trained using Muse Wonder, bringing Active Memory and Active Focus to all Muse and Meta AI users (Frederic Lardinois/The New Stack)aifleet-techmeme
How we got here
The New Stack reported that Meta developed these changes by training new models using the Muse Wonder frontier model.reported
Evidence: reported
Meta introduces specialized versions of its existing models, trained using Muse Wonder, bringing Active Memory and Active Focus to all Muse and Meta AI users (Frederic Lardinois/The New Stack)aifleet-techmeme
What is new
The primary change is the availability of Active Memory and Active Focus features. These capabilities are now shipped to all Muse and Meta AI users through these newly trained specialised model versions.
What to watch
Developers should note that while the features are broadly available, the underlying Muse Wonder model itself is not. Meta is keeping the frontier Wonder model reserved for harder subscriber tasks.
Availability and limits
Access to Active Memory and Active Focus is now universal for Muse and Meta AI users. However, the source material indicates a limitation: the Muse Wonder model used to train these versions is retained for harder subscriber tasks.
Why it matters
- Users of Meta AI and Muse now have access to Active Memory and Focus features.
- The Muse Wonder model is reserved for harder subscriber tasks, distinguishing it from the general release.
- The deployment brings these specific capabilities to all users of the Meta AI and Muse platforms.
Topics: Meta, AI models, Muse, Active Memory, Active Focus
Updated
The strongest doubt
Counterpoint from the sources
The source material does not specify if there are any performance degradations in the specialised models compared to the original Muse Wonder system.
Ask the article
Local keyword search of this article; no generated answers.Did this article give you the evidence you needed?
Your choice stays on this device. No public tally.Terms in this story
Active Memory
A feature allowing the AI to retain context from previous interactions.
Active Focus
A capability enabling the AI to prioritise specific information during processing.
Muse Wonder
Meta's frontier AI model used for training new versions and complex tasks.
Questions readers ask
Do all users get access to the Muse Wonder model?
No, Meta is keeping the Muse Wonder model for harder subscriber tasks, though all users get the features derived from it.
What new features are available to all users?
Active Memory and Active Focus are now available to all Muse and Meta AI users.
How were the new models created?
Meta trained new versions of its existing models using the Muse Wonder frontier model.
Connected stories
Reading receipt
- Meta trained new versions of existing models using its frontier Muse Wonder system.
- Active Memory and Active Focus features are now available to all Muse and Meta AI users.
- The underlying Muse Wonder model remains reserved for harder subscriber tasks.