Artificial intelligence systems based on neural networks—such as ChatGPT, Claude, DeepSeek or Gemini—are extraordinarily powerful, yet their internal workings remain largely a "black box." To better ...

Researchers use statistical physics and "toy models" to explain how neural networks avoid overfitting and stabilize learning in high-dimensional spaces.

The observational track of Typhoon "Danas" (solid line) along with forecasted paths (dashed lines) depicted on the FY-4B satellite visible light imagery at 08:00 BST on July 6, 2025. The dashed lines ...

For decades, simulation has been the cornerstone of scientific discovery and hardware design. But despite the advances we’ve made in computing power and tooling, the way we run simulations hasn’t ...

Harvard University physicists have developed a simplified, physics-inspired mathematical model to better understand how neural networks learn. Published in the Journal of Statistical Mechanics, the ...

But as the chip maker seeks to parlay its semiconductor dominance into control of the full stack, vendor lock-in is a major risk, says one analyst. Nvidia hopes that its new open-source AI model for ...

Research from Los Alamos National Laboratory shows that a new rock physics model will provide more comprehensive and actionable data about how carbon dioxide changes rock properties throughout ...