🔬 AI Research

Academic papers, research breakthroughs, and scientific advances in machine learning and AI.

585 articles

Flawed Kaggle Datasets Used in Clinical Models Spark Retractions

Researchers found celebrity images and thousands of duplicated entries in datasets that were used to create predictive models, including some in clinical... Source: the-scientist.com

Survey Bridges Evolutionary Algorithms and Reinforcement Learning

A comprehensive survey titled "Bridging Evolutionary Algorithms and Reinforcement Learning: A Comprehensive Survey on Hybrid Algorithms" appears on arXiv as... Source: letsdatascience.com

Gemini for Science: AI experiments and tools for a new era of discovery

Gemini for Science is a new collection of science tools and experiments to expand the scale and precision of scientific exploration. Source: blog.google

Quantum Computers Maintain Learning Edge Despite Real-World Noise

Thirty to forty qubits is all it took to reveal a clear advantage for quantum machine learning, a threshold previously thought far beyond reach. Source: quantumzeitgeist.com

Machine Learning Personalizes Depression Treatment with the Help of Wearable Technology

UC San Diego researchers report substantial reductions in symptoms of depression by using machine learning to tailor behavioral interventions based on data... Source: today.ucsd.edu

Machine Learning Doubles Depression Remission Rate

Summary: A pioneering, first-of-its-kind study demonstrated that a personalized, machine-learning-guided lifestyle coaching program can nearly double the... Source: neurosciencenews.com

Researchers improve AI's ability to learn new tasks without sacrificing performance

A new framework allows AI models that have already been trained to learn new tasks without sacrificing performance when performing old tasks. Source: techxplore.com

This week in AI research: Fields medalist says GPT-5.5 Pro did PhD-level math in an hour, Anthropic teaches Claude to 'dream'

The past week or so, Anthropic's gated Mythos model remains a news fixture, drawing White House policy attention, a Mozilla proof point and a somewhat... Source: rdworldonline.com

Machine learning proves that graphene is hydrophobic

For more than a decade, a fundamental mystery has surrounded graphene—the one-atom-thick "wonder material" known for its exceptional strength, conductivity,... Source: phys.org

Scaling and quantization of a foundational deep learning model for network biology

We developed a Geneformer model with an expanded pretraining dataset of more than 100 million single-cell human transcriptomes. The increased data diversity... Source: nature.com

Apple shares recordings and research from recent privacy-focused AI and ML workshop

Apple has published four recordings and a research recap from its 2026 Workshop on Privacy-Preserving Machine Learning & AI. Source: 9to5mac.com

Amsterdam researchers have built a material that learns without software, and its moving structure remembers past shapes as if intelligence were embedded in the object itself

Most materials are predictable. Push them, bend them, heat them, and they respond the same way every time. Robots are different, but they usually need a... Source: ecoticias.com

Disney Research Releases ReActor Motion-Retargeting System

Disney Research published a paper and demo video describing **ReActor: Reinforcement Learning for Physics-Aware Motion Retargeting**, a system that adapts... Source: letsdatascience.com

New work by Turing Award winner Sutton: Solve a major flaw in streaming reinforcement learning with a 1967 formula

At the end of 2024, a paper titled "Streaming Deep Reinforcement Learning Finally Works" (arXiv:2410.14606) sparked extensive discussions in the academic... Source: eu.36kr.com

Why Massive AI Models Actually Generalize Better

Researchers use statistical physics and "toy models" to explain how neural networks avoid overfitting and stabilize learning in high-dimensional spaces. Source: neurosciencenews.com

RVPO: Risk-Sensitive Alignment via Variance Regularization

Current critic-less RLHF methods aggregate multi-objective rewards via an arithmetic mean, leaving them vulnerable to constraint neglect:… Source: machinelearning.apple.com

A Major Paper Claiming AI Is Good for Students Just Got Retracted, Which Is Very Bad News for Advocates of AI in the Classroom

The publishers of the journal Nature retracted a much-touted study that claimed AI had a "large positive impact" on learning. Source: futurism.com

Normalizing Flows with Iterative Denoising

Normalizing Flows (NFs) are a classical family of likelihood-based methods that have received revived attention. Recent efforts such as… Source: machinelearning.apple.com

Revolutionizing clinical trials with machine learning

image: Design Framework for Adaptive Randomization with Interim Analyses. view more. Credit: Y.Park and S.Nycklemoe, “MARGO: Machine Learning-Assisted... Source: eurekalert.org

Machine learning identifies five distinct Parkinson’s disease subtypes

A new study from VIB and KU Leuven has revealed that Parkinson's disease comprises five distinct molecular subtypes, each requiring tailored therapeutic... Source: drugtargetreview.com