Researchers have developed a unified variational autoencoder framework that effectively disentangles latent representations without relying on pre-defined ground truth factors.
A new benchmark dataset reveals that even advanced artificial intelligence struggles with the visual reasoning required to solve complex materials science problems.
New research reveals a surprisingly simple principle for optimizing AI performance: drastically reduce the size of the initial ‘draft’ model used in speculative decoding.
A new deep learning approach accurately models nonlinear gas flow through porous media, overcoming limitations of traditional methods.
New research reveals that standard machine learning techniques can surprisingly provide enough exploration to effectively solve contextual bandit problems, simplifying algorithm design.
New research demonstrates how deep neural networks can achieve state-of-the-art performance in predicting outcomes from data where observations aren’t independent.
A new approach leverages coordinated drone swarms and artificial intelligence to rapidly and accurately locate the origins of harmful chemical emissions.
Researchers have developed a system that refines AI-generated speech in real-time using feedback from powerful audio models, dramatically improving quality and naturalness.
New research demonstrates that models inspired by biochemical reactions can match or outperform spiking neural networks in supervised learning tasks.
Researchers have developed a new AI workflow that marries the accuracy of traditional fonts with the creative freedom of image generation to produce remarkably precise and stylish text within images.