Leaked Astra and Fable sizes are far below 10T, says X user: scale barely maps to capability now
teortaxesTex · x · 2026-09-14
X user teortaxesTex riffed on @scaling01's reaction ("Astra is much smaller than I thought — looped transformers are going to eat the world"), noting that neither Astra nor Fable is anywhere near "10T" parameters and that people should stop getting worked up over speculative model sizes, since scale has become very tenuously related to capability.
The exchange hints that rumored next-gen models may use looped transformer architectures with far smaller parameter counts than assumed.
More from Models
- Playing GeoGuessr against DeepSeek 4.1 Flash: it's fairly good at geolocation — PMinervini · 2026-09-14
- DeepSeek V4.1 benches low but feels top-tier at 4x speed, insiders say — teortaxesTex · 2026-09-14
- Best small LLMs for writing on 4GB VRAM? Reddit thread weighs the options — Mysterious-Comment94 · 2026-09-14
- Free ChatGPT solves a decade-old maths dice problem in 13 minutes — Chris_Armstrong · 2026-09-14
- David Bellamy clarifies his experiment used K2 Horizon, an open-weights 375B LLM — JeremyNguyenPhD · 2026-09-14
- Reddit user gets ChatGPT 'Cyber Abuse' warning, appeal denied in an hour with no explanation — TheBeaconCrafter · 2026-09-14