@Historionome@axvai Ca serait pas très malin, et les conséquences seraient désastreuses pour la France. Mais bon, on est habitué aux Français qui font n'importe quoi.
@miracle23maker@skara_io@ylecun I will never bet against a french guy. Highest field medal per capita by a wide margin (small nations are close but its because its like 1 guy won it). And AI is a math problem.
Excited to see S1 finally go public! 🚀 This release also marks exactly one year since I transitioned from vision research to robotics. If I had to share the single biggest lesson from that journey, it’s this:
Before asking "How do we train a better model?", we need to ask "How should the model behave at inference time?"
In academia, we tend to attack problems the same way: set up a benchmark, tweak the architecture or training recipe, and watch the score climb. Today, much of robotics research runs this similar loop: pit VLAs against WAMs, introduce a new loss function, test a different action representation. Model performance on benchmarks keeps improving, but inference-time behavior stays simplistic and fixed.
But my past year at @SkildAI taught me that raw model performance isn't the ultimate driver of real-world success. Inference-time behavior is.
The real world guarantees high stochasticity and endless out-of-distribution scenarios. The dominant deployment paradigm (taking the past X seconds of observations to predict the next Y seconds of actions) won't get us to general-purpose robotics, because every robotic action has compounding consequences. A vision model can guess wrong on a static test set with zero impact on the next sample. A robot cannot. It operates in a closed loop: its own mistakes create the very distribution shifts it must then survive. "Collect more data, fine-tune, redeploy" is not a scalable answer. To truly generalize, a model must adapt on the fly.
So the order of operations has to invert. First, define an inference-time behavior scalable and robust enough to handle the real world. Then work backward to shape data collection and model training. Everything in the S1 release stems from this inversion.
Language modeling already taught us this lesson: LLMs didn't crack complex multi-step math just because someone threw more tokens at pre-training. Instead, the breakthrough came from changing the inference-time behavior first (reasoning step by step) and then redesigning training around it. Betting that data scale alone will magically produce emergent physical skills is just as naive. Robotics needs the same shift.
S1 is proof that working backward from real-world inference changes everything. And this is just the starting point.
Check out the release below, and enjoy the rest of your day! 👇
@dragosevacsutra@arnesa_kustura Bosnian Serbs or Bosnian Croats are as much Serbs or Croats as Jesus was a Jew.
Well, actually there is way more supporting facts to call Jesus a Jew than to call a Bosnian-Croat/Serb a Croat or a Serb.
@ylecun@ziv_ravid Yes because in the domain of prediction I think the horizon of visibility displays a synthesis of the prereflective experience and of the achievements of reflection insofar as they preserve ontological continuity with the visible.
LLMs can't synthetize a prereflection.
How do we build sparsity into JEPA representations by design, while preserving task-relevant information?
Introducing Rectified LpJEPA, a JEPA architecture that learns sparse, non-negative, informative representations through principled distributional regularization. 📐
📄 Paper: https://t.co/CZtKHVvKT4
💻 Code: https://t.co/xIJMYkK9fT
📝 Blog: https://t.co/OFCl3FAmik
(1/n)
@ketansingh279@wyqtor@jm_alexia@ilyasut@ylecun You a fucking dumbass. Yann was NEVER cynic on LLM practicality. Yann is an AI researcher whom main objective is to solve AGI. He said LLMs are very useful, but they won't lead to AGI.
@neu_rips@giffmana " A joint isotropic gaussian has the property that marginals along all projections are also gaussian (and vice versa).
Not true of Laplace."
Yann LeCun
@hugofrns Ta gueule le mulâtre.
De Gaulle : " Nous sommes avant tout un peuple européen de RACE BLANCHE, de culture grecque et latine et de religion chrétienne. "
Tu seras jamais Français
@hugofrns@NNacoub Mdrr qu'est-ce que tu juges avec ta gueule de malgache et bougnoule.
De Gaulle a été clair " Nous sommes avant tout un peuple européen de RACE BLANCHE, de culture grecque et latine et de religion chrétienne. "
T'es pas blanc toi.
@SurfingCelia@conileetmax@franceinfo Oui énormément de drame le 7 Octobre de la part du Hamas. Puis énormément de drame de la part d'Israel, le 8 Octobre, le 9 Octobre. le 10,11,12,13,14,15,16..Novembre,Decembre, Janvier,Fevrier, Mars, Avril, Mai... plus de 18000 enfants.
Mais c'est vrai il n'y a que le 7 Octobre 😕
@Dechandelet@Historionome Réfléchissez deux minutes. Si les Russes frappent en UE, ça sera logiquement à l'Ouest vu que c'est de là que vienne la majorité de l'aide.