Nature

Information Bottleneck in Deep Learning and Neural Networks

The information bottleneck (IB) principle is a powerful information‐theoretic framework that seeks to compress data representations while preserving the information most pertinent to a given task.
Phys.org

Q&A: How mathematics can reveal the depth of deep learning AI

Artificial intelligence (AI) is increasingly prevalent, integrated into phone apps, search engines and social media platforms as well as supporting myriad research applications. Of particular interest ...
Phys.org

Using geometry and physics to explain feature learning in deep neural networks

Deep neural networks (DNNs), the machine learning algorithms underpinning the functioning of large language models (LLMs) and other artificial intelligence (AI) models, learn to make accurate ...
Semiconductor Engineering

SpiNNaker2 Neuromorphic Platform: HW-Aware Fine-Tuning of Spiking Q-Networks (TU Dresden Et Al.)

A new technical paper titled “Hardware-Aware Fine-Tuning of Spiking Q-Networks on the SpiNNaker2 Neuromorphic Platform” was published by researchers at TU Dresden, ScaDS.AI and Centre for Tactile ...
Ars Technica

How a stubborn computer scientist accidentally launched the deep learning boom

During my first semester as a computer science graduate student at Princeton, I took COS 402: Artificial Intelligence. Toward the end of the semester, there was a lecture about neural networks. This ...
The Conversation

Physics Nobel awarded to neural network pioneers who laid foundations for AI

The 2024 Nobel Prize in Physics has been awarded to scientists John Hopfield and Geoffrey Hinton “for foundational discoveries and inventions that enable machine learning with artificial neural ...
Quanta Magazine

Novel Architecture Makes Neural Networks More Understandable

“Neural networks are currently the most powerful tools in artificial intelligence,” said Sebastian Wetzel, a researcher at the Perimeter Institute for Theoretical Physics. “When we scale them up to ...