Perry Dong @perryadong · 30 Jul 2026
We analyze the performance of pretraining against training from scratch, and show that naive Q-function pretraining does not accelerate fine-tuning (2/6) https://t.co/iC77krarpl
3 074Views
10Likes
1Reposts
1Replies
0Quotes
2Bookmarks
Is that a lot?
1.73×vs this author's median1 780 views is typical
83Percentile for this authorof 6 recent posts
0.43×vs under 10K median7 156 views is typical
2.7× audienceReachviews ÷ followers
0.39%Engagement rateof viewers reacted