3 ms·
Correct me if I'm wrong but LeCun is focused on learning from video, whereas Fei-Fei Li is doing robotic simulations. Also I think Fei-Fei Li's approach is stil
by dauertewigkeit 11mo ago
Correct me if I'm wrong but LeCun is focused on learning from video, whereas Fei-Fei Li is doing robotic simulations. Also I think Fei-Fei Li's approach is still using transformers and not buying into JEPA.
- Legend2440 11mo agoJEPA is not an alternative to transformers, it is built out of transformers.
- HarHarVeryFunny 11mo agoJEPA is trying to learn real world regularities by predicting internal perceptual representations (using a transformer). It seems a good direction to go in, but tbh he doesn't seem to have taken it very far, and it is the transformer doing all the heavy lifting.