
berry
872 posts




Yann LeCun and his team can't stop cooking "LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels" One of the biggest bottlenecks of JEPA is they are hard to train, and this new research changes that. They propose LeWorldModel, which shows that a small model can learn a usable world model directly from raw pixels end-to-end. Sitting at 15M parameters, they made it without needing heuristics and avoiding anti-collapse hacks while staying competitive and planning up to 48x faster. Making JEPA based modeling much more accessible, cheaper, and stabler.





just bought a lenovo legion 5 goodbye macbook suffering! >would you be interested if i stream setting it up?



Grok 4.20 is now officially out of Beta. It's now on Auto, Fast, Expert & Heavy.

















