tweetindex

Sean McLeish @SeanMcleish · 11 Nov 2025

Looped latent reasoning models like TRM, HRM, Ouro and Huginn are great for reasoning, but they’re inefficient to train at larger scales. We fix this by post training regular language models into looped models, achieving higher accuracy on a per training FLOP basis. 📜1/7 https://t.co/mXVIDhtHBz
65 922Views
392Likes
65Reposts
9Replies
13Quotes
293Bookmarks

Is that a lot?

17.3×vs under 10K median3 808 views is typical
100.0× audienceReachviews ÷ followers
0.73%Engagement rateof viewers reacted

Compare with the benchmark table →

Open on X →