Aray Karjauv·19h agoWhat Does a Vision Model See Without Labels?How activation maps and relevance propagation can shed light on deep models without any labels.
Hola Adrakey, PhD·Nov 6When Preventing Collapse Destroys Useful Representations: The Hidden Role of Optimisation in LeJEPAAn experimental journey into LeJEPA, Gaussian regularisation, and the difference between geometric diversity and discriminative…
Eleventh Hour Enthusiast·Nov 5Proving and Improving the Platonic Representation Hypothesis with JEPAsPaper Review
Mohammed El Amine Mokhtari·Nov 1LoRe-MIM: Pretraining Whole-Body PET/CT on Unlabeled ScansA masked-image pretext that swaps PET and CT blocks helps a SwinUNETR find liver and bone lesions and drop false calls.
InAccredianbyAdiNagare·Oct 23Representation Learning: The Art of Learning What MattersTeaching Machines to be Self- Reliant!
InThe Singularity ProjectbySteven Vincent·Oct 22Yann LeCun’s Joint Embedding Predictive Architecture (JEPA) and the General Theory of IntelligenceIs JEPA a new architecture or an extension of existing technologies?
Maneaditya·Sep 28What If AI Didn’t Need to Recreate the World to Understand It? Inside JEPAImagine showing an AI a video of a ball being thrown into the air.
Jorge Jonas Rodríguez Godoy·Sep 21Solo las personas que asumen las consecuencias de sus actos pueden sentirse merecedoras de sus…Una reflexión original de Jorge Jonas Rodríguez Godoy © 2026
Nicola Rohrseitz·Sep 17How AI Learned to Learn from the InternetLanguage already contained meaningful relationships. Self-supervised learning turned them into opportunities to learn
Aditya Randive·Sep 15The Cave and the Coin FlipPlato v/s Aristotle — Which one is right about intelligence?A response icon1A response icon1
