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

Stop throwing money at GPUs for unoptimized models; using smart shortcuts like fine-tuning and quantization can slash your ...

Chinese artificial intelligence lab Moonshot AI has raised $2 billion in funding at a valuation exceeding $20 billion.

Statistical principles show you don’t need a nefarious plot to explain clusters of missing scientists and lab workers ...

A leaf may appear to be one of the simplest structures in nature, thin, delicate and easily overlooked. At first glance, it ...

Penn researchers have developed a smarter AI method for solving notoriously difficult inverse equations, which help ...

Harvard University physicists have developed a simplified, physics-based mathematical model to better understand how neural ...

Spiking Neural Networks (SNNs) represent the "third generation" of neural models, capturing the discrete, asynchronous, and energy-efficient nature of ...

Harvard University physicists have created a simplified mathematical model to study how neural networks learn, using statistical physics to uncover underlying patterns. The approach, likened to early ...

Artificial intelligence systems based on neural networks—such as ChatGPT, Claude, DeepSeek or Gemini—are extraordinarily ...

The first major fruits of the x86 Ecosystem Advisory Group (EAG) have come in the form of ACE, a new set of matrix ...

Penn Engineers have developed a new way to use AI to solve inverse partial differential equations (PDEs), a particularly ...