Science

Researchers are moving beyond superficial visual cues to identify deepfakes by focusing on subtle inconsistencies in facial movements and temporal artifacts.

A new framework combines the power of graph neural networks with pattern analysis to build more transparent and reliable graph-based artificial intelligence.

A new study pits traditional feature engineering against deep learning to determine the best approach for automatically assessing the condition of valuable Saint-Gaudens Double Eagle coins.

A new framework, Arbitrage, dramatically improves the efficiency of large language models by intelligently switching between draft and refined text generation.

New theoretical work provides a framework for understanding and predicting the sample complexity of deep learning models, offering insights into how feature learning impacts generalization.

New research explores how artificial intelligence agents can thrive in competitive labor markets by strategically enhancing their skills and building reputations.

A new approach to enhancing training data is delivering significant improvements in perceptual quality and restoration fidelity.

New research tackles the challenge of reliably assessing advertising policy performance in real-world auctions where market prices aren’t fixed.

New research shows that advanced AI models can detect faint gravitational waves directly from real-world data, reducing reliance on massive simulations.

A new approach to updating neural network weights based on data sensitivity dramatically improves generalization to unseen data, offering promising results in fields like climate modeling.