UC Berkeley Cuts Transformer Memory Usage With New Attention
UC Berkeley researchers unveil a novel attention mechanism that dramatically reduces memory consumption in Transformer m…
Latest articles in Research
UC Berkeley researchers unveil a novel attention mechanism that dramatically reduces memory consumption in Transformer m…
Carnegie Mellon researchers unveil a new continuous RL framework that dramatically improves robotic learning efficiency …
New OpenAI research shows large language models develop internal planning mechanisms without explicit training, challeng…
New MIT research shows AI tools boost productivity for knowledge workers without eliminating positions, challenging wide…
Meta's FAIR lab releases SA3, bringing real-time 3D scene understanding to its open-source computer vision framework.
Google DeepMind unveils AlphaFold 4, achieving atomic-level accuracy in modeling how drugs bind to proteins.
Stanford's 2025 AI Index Report reveals AI job displacement remains far below early predictions, though workforce transf…
New research reveals that just 10 minutes of AI assistant use can measurably reduce cognitive effort and problem-solving…
MIT researchers unveil a graph neural network approach that achieves state-of-the-art accuracy in protein structure pred…
Google DeepMind achieves state-of-the-art results in protein structure prediction, accelerating drug discovery and biolo…
A new open-source project claims to detect patterns in chaotic systems like stock markets without massive parameter scal…
Japanese AI firm Preferred Networks unveils breakthrough in massively parallel robot simulation, enabling thousands of v…