Distillation is over btw not because Fable access is pulled you just won't get far "distilling" this thing via SFT would be a kickass reward model though
10:10 PM · Jun 12, 2026 · 2K Views
Teortaxes argues supervised fine-tuning is no longer effective for distilling advanced models like Fable
They suggest using Fable as an RL reward model.
143052.1K
Sentiment
Sentiment building, check back later.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS317LIKES6
kalomaze@kalomaze
@teortaxesTex Policy Gradient Believer
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex
Distillation is over btw not because Fable access is pulled you just won't get far "distilling" this thing via SFT would be a kickass reward model though
1hViews 317Likes 6Bookmarks 0