Video Joint Embedding Predictive Architectures (V-JEPA) learn generalizable off-the-shelf video representation by predicting masked regions in latent space with an exponential moving average (EMA)-updated teacher. While EMA prevents representation collapse, it complicates scalable model selection and couples teacher and student architectures. We revisit masked-latent prediction and show that a frozen…
A new chip design from UC San Diego could make data centers far more energy-efficient by rethinking how power is converted for GPUs. By combining vibrating piezoelectric components with a clever circuit layout, the system overcomes limitations of traditional designs. The prototype achieved impressive efficiency and delivered much more power…
Researchers at Tsinghua University developed the Optical Feature Extraction Engine (OFE2), an optical engine that processes data at 12.5 GHz using light rather than electricity. Its integrated diffraction and data preparation modules enable unprecedented speed and efficiency for AI tasks. Demonstrations in imaging and trading showed improved accuracy, lower latency,…