Sakana AI @SakanaAILabs · 03 Sep 2026
Can a strong general-purpose vision model distinguish real from AI-generated images using only a simple decision rule on frozen representations? Our latest results, to be presented at #ECCV2026, show it can. Blog: https://t.co/VtFAm5kIKy Paper: https://t.co/9Q84QYhkum New image https://t.co/hSd9oqYVDD
13 495Views
119Likes
14Reposts
2Replies
3Quotes
50Bookmarks
Is that a lot?
0.96×vs this author's median14 104 views is typical
33Percentile for this authorof 12 recent posts
3.19×vs 100K–1M median4 224 views is typical
9.73%Reachviews ÷ followers
1.02%Engagement rateof viewers reacted