Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
A new method called OmniHarness applies symbolic policy learning to unified multimodal large language models (MLLMs) and multi-agent systems to make visual generation more generalizable. The work identifies three limitat…
Aito published beta benchmarks claiming its predictive database matches or beats tuned ML baselines on high-cardinality tasks with zero training time, including 68.5% accuracy on invoice routing acceptor versus FLAML's 5…
Lightstorm Vision, the company founded by filmmaker James Cameron, has acquired Outsyders, an AI and machine-learning 3D production and post-production technology company founded in 2024 by Paul Becker, Jared Sandrew and…
Researchers submitted a paper to arXiv on 15 Sep 2026 proposing Episode-Normalized Conformal Prediction (ENCP), a method that rescales a nonconformity score by a policy's residual confidence and calibrates one maximum sc…
Researchers released ScienceBuddy, an interactive scientific research workspace built on a paradigm they call recursive-in-recursive self-improvement, submitted to arXiv on 15 Sep 2026. The system couples harness evoluti…
Roboflow's MIT-licensed Supervision computer vision library crossed 50,000 GitHub stars and claims over 1 million PyPI downloads per month, following its 0.30.3 release on September 14, the fourth release in six weeks. T…
An engineer published a technical due-diligence checklist for evaluating ML-based hiring tools, urging buyers to verify per-decision explanations, training-data provenance, and independent bias audits with four-fifths ad…
Researchers developed ASDint, a spiking neural network model, paired with SyNC, an ASIC-based hardware architecture, to analyze glutamatergic neuron synapse dysfunction in Autism Spectrum Disorders and related neurologic…
Google Research introduced R4T, an RL-trained diffusion retriever that optimizes fan-out query sets for diversity and coverage, achieving 12x to 20x latency speedups over autoregressive language models. The 53.9M-paramet…
A developer outlined a multi-tier strategy for cutting AI latency in mission-critical, real-time enterprise applications, combining semantic caching, streaming responses, and routing time-sensitive tasks to smaller, hard…
TypeSafe launched Jev, a "System One Model" that only decides, classifies, routes, and scores, claiming it is over 100x faster and over 200x cheaper than small frontier LLMs, and the announcement sat atop Hacker News all…
Airbus AI Research initiated scikit-decide, an open-source AI framework for reinforcement learning, automated planning and scheduling that auto-matches decision-making problems with compatible solvers. The framework rece…
Researchers at The University of Manchester, Shandong Jiaotong University and Harbin Engineering University developed a machine learning framework that predicts the thermal response of three-sided protected steel beams d…
Graduate fellows Priyamvada Nambrath and Eleanor Webb used the open-source Handwritten Text Recognition platform eScriptorium, which interfaces with the Kraken automatic text recognition system, to transcribe premodern m…
Prior Labs released TabPFN-3.5, a tabular foundation model that scores 0.375 on the private leaderboard of the 2015 Otto Group Product Classification Challenge, beating the winning solution's 0.382, according to Prior La…
A developer published a beginner's guide explaining unsupervised learning, the branch of machine learning in which algorithms discover patterns and groups in data without labeled answers. The guide contrasts it with supe…
A developer outlined a conceptual approach to configuring smart traffic systems, framing the problem as a distributed control loop that uses sensors, data analytics, and AI/ML techniques such as reinforcement learning an…
A developer's analysis of realistic machine learning learning timelines finds that the popular "job-ready in 6 weeks" promise is disconnected from reality for nearly everyone. The breakdown estimates 9 to 14 months of co…
Korea Hydro & Nuclear Power (KHNP) won the Technology Transfer Award from the Electric Power Research Institute (EPRI) for a joint wind power performance improvement project in which KHNP developed machine learning techn…
A developer community analysis argues that 2026's leading large language models are deliberately trained to minimize stored factual knowledge in favor of compact, reusable reasoning procedures, citing Simon Willison's te…