Ant Group's Robbyant has open-sourced LingBot-Vision, a 1 billion parameter vision foundation model that uses masked...
Jesse Thaler, a physics professor and director of the NSF AI Institute for Artificial Intelligence and Fundamental...
The article discusses Fables writing GPU kernels, AI automation capabilities, and analog computation approaches. These...
Baidu has developed 'Unlimited OCR,' an optical character recognition system that can process dozens of document pages...
A large-scale study of 26,000 Chinese students reveals that while AI users complete homework faster and score higher...
GeneBench-Pro is a new benchmark designed to evaluate AI performance in genomics, biology, and scientific research...
An inaugural Music Technology Research Showcase celebrates the work of students in a new graduate program, featuring a...
Princeton University researchers created CEO-Bench, a benchmark test where AI agents manage a simulated software...
David Autor, a faculty member since 1999 and leading researcher in artificial intelligence and the future of work, has...
Insurers are adopting generative AI and diffusion models for catastrophe modeling to generate synthetic weather events...
UC San Diego's DFlash introduces a lightweight block diffusion model for speculative decoding that drafts entire token...
This tutorial demonstrates how to implement NVIDIA's Canary-1B-v2 model for multilingual automatic speech recognition,...
Researchers have developed a new chip that combines an efficient algorithm with specialized hardware to enable tiny...
MoonMath AI has open-sourced a HIP attention kernel for AMD MI300X processors that outperforms AMD's official AITER v3...
The article compares three KV cache compression techniques—TurboQuant, OSCAR, and EpiCache—that address the memory...
Researchers have discovered that a previously overlooked class of algorithms performs significantly better than...
The article discusses the challenge of collecting training data for physical AI systems, noting that while large...
Researchers have developed a new spatial memory system for robots that efficiently captures and recalls details about...
This article demonstrates how to implement xFormers, a toolkit for building memory-efficient Transformer models on GPUs...