Adrien Bardes
@AdrienBardes
PhD Student at @MetaAI & @Inria with Yann LeCun and Jean Ponce, interested in self-supervised learning and computer vision.
ID:787639668
http://adrien987k.github.io 28-08-2012 19:01:18
42 Tweets
538 Followers
231 Following
OpenAI may secretly know that you trained on GPT outputs!
In our work 'Watermarking Makes Language Models Radioactive', we show that training on watermarked text can be easily spotted ☢️
Paper: arxiv.org/abs/2402.14904
Pierre Fernandez AI at Meta École polytechnique Inria
Bag of Image Patch Embedding Behind the Success of Self-Supervised Learning
Yubei Chen, Adrien Bardes, @LiZengy, Yann LeCun
tl;dr: bag of patches and compositional structure of the horse is all you need.
(except where is Dinov2 comparison?) and
openreview.net/pdf?id=r06xREo…