User Flags F5.1 Silently Reverting to Opus 5/4.8 on Long Overnight Proof Tasks
DimitrisPapail · x · 2026-09-06
Dimitris-Papail says his biggest — perhaps only — complaint about F5.1 is that it silently reverts to Opus 5 or 4.8 while you're asleep, for random reasons, even on silly information-theory math problems that need to run for days due to a computerized proof process.
The complaint highlights a real usability issue: long-running tasks can end up on a different model version than expected, without notice.
More from Models
- Testing AI models on games is a mess: famous games leak guides, obscure games bore everyone — banteg · 2026-09-06
- Should Models Be Trained on Certain Games? Models May Be 'Too Smart' to Play — scaling01 · 2026-09-06
- 'LLM Psychosis' Reframed: Models Mistaking Their Own Thoughts for Injected Ones — paul_cal · 2026-09-06
- Ethan Mollick: fable 5.1 and astra exceed routine knowledge work — expect amplification — emax · 2026-09-06
- GPT-6 Astra users hit frequent 'model at capacity' errors — bytebot · 2026-09-06
- Astra User Says Weekly Token Limits Too Tight: 24-Hour Run Still Unfinished — Wooden_Drag9473 · 2026-09-06