Meta retrains its everyday models on flagship Muse Wonder: 2.4x less forgetting, +67% task completion
xiaohu · x · 2026-11-09
Meta didn't ship its flagship Muse Wonder directly into every chat. Instead it used Muse Wonder as a training source to retrain the models behind Muse and Meta AI into specialized versions, now rolling out to everyday users, targeting two pain points: forgetting context in long chats and stalling mid-task.
Active Memory: trained to distinguish temporary vs. long-term info, reason across time, and detect stale facts. Accuracy on context-critical tasks rose from 76% to 90%, with missed critical context dropping from 24% to 10% (the "2.4x less forgetting" claim).
Active Focus: on errors or unresponsive tools it retries, recovers, or reroutes; completion on disrupted everyday tasks rose 67%. In one demo, when the user switched to "follow up in writing," Muse pivoted to emailing an insurance claim without losing prior work.
Muse's Personal Superintelligence Index score rose from 38% to 47%. Muse skews toward agentic execution; Meta AI toward collaborative thinking.
More from Models
- User says Terra Max's batch browser use is the main reason for their ChatGPT subscription — amplifiedamp · 2026-11-09
- DeepSeek V4.1 Flash 1017 hits 1,000 tok/s on Vercel while a home 4x 3090 rig does 65 — AccBalanced · 2026-11-09
- Fastino open-sources GLiDE decision models: 27B scores 72.49 on Decision Index at 112ms — RobertTLange · 2026-11-09
- Grok usage limits reset within hours after Musk-led engineering review — XFreeze · 2026-11-09
- Qwen4-27B leaks questioned: 480 tps benchmark may be Qwen4-Max-preview on low batches — teortaxesTex · 2026-11-09
- The 'Jensen loves Chinese models' joke: not the LLMs everyone assumed — haider1 · 2026-11-09