SDNist are intended as tools to encourage investigation and discussion of deiden-tification algorithms, and they are not intended or suitable for product evaluation. The National Institute of Standards and Technology does not endorse any algorithm included in these resources.

Deidentification algorithms are vulnerable to the same bias and privacy issues that impact other data analytics and machine learning applications, and it can even amplify those issues by contaminating downstream applications.

Roland Martin defends Rep. Jasmine Crockett against white progressive critics of her Texas Senate bid, telling them to "sit their a-- down" on Tuesday's episode of "At Our Table."

Wing's drone deliveries are coming to 150 more Walmarts The service expansion will reach Walmart customers in Los Angeles, St. Louis, Cincinnati, Miami and other US metro areas. Don't be surprised if you see even more drones delivering groceries across the US since the Alphabet-owned Wing announced another service expansion with Walmart over the next year. The partnership said that drone delivery services will be available at 150 more Walmart locations in Los Angeles, St. Louis, Cincinnati, Miami and more metros that have yet to be announced. According to Wing, its top 25 percent of customers have ordered its delivery drones up to three times a week. To meet growing demand, Wing and Walmart said it will serve up to 40 million US customers and build up a network of 270 delivery locations by 2027.

American Airlines' AI system delays flights to help passengers make connections, expanding testing from Dallas-Fort Worth to Los Angeles, Charlotte, Miami and elsewhere.

We may earn revenue from the products available on this page and participate in affiliate programs. At, we've published our prestigious Best of What's New list since 1988. For 153 years, we've celebrated the science and technology that shapes our everyday lives and launches humanity forward. Innovation doesn't follow a straight path, and the detours, stumbles, and dead ends force great minds to pioneer change. Looking back at the early days of our Best of What's New lists, we see technologies that now seem quaint or have been completely forgotten, but we also see the roots of future greatness. Our list this year is the culmination of countless hours of debate, hands-on testing, and expert conversations. This is the Best of What's New 2025. From the most detailed movie of the night sky ever made to the first commercial soft landing on the moon, this year has been an inflection point for exploring and understanding the vast expanse above our heads. We also saw breakthroughs in small changes to commercial airliners that improve efficiency, as well as a new type of rocket engine that might be the future of extremely high speed air travel, plus the closest view of Mercury we've ever seen! Vera C. Rubin Observatory by U.S. National Science Foundation & Department of Energy: World's largest digital camera to conduct 10-year survey of the night sky Prepare to see space like never before. The Vera C. Rubin Observatory is a groundbreaking US-funded project that will capture the most detailed, dynamic map of the night sky ever made. Using the world's largest digital camera, it will capture a time-lapse of the entire sky every few nights to reveal billions of objects and catch fast-changing events like supernovae and near-Earth asteroids. Its massive dataset will help scientists better understand dark matter, dark energy, and the structure of the universe while also improving planetary defense. The 3,200-megapixel Legacy Survey of Space and Time (LSST) camera is the size of a small car and twice as heavy, tipping the scales at 6,000 pounds. The sensor's huge number of megapixels is equivalent to 260 modern cell phone sensors. The camera is so powerful, it could snap a clear image of a golf ball from 15 miles away. By making its data widely available, the observatory will also open new doors for discovery for researchers, students, and citizen scientists around the world. Deployed on Boeing 787-9 aircraft starting in January, the coating uses tiny, sharkskin-like grooves called riblets to guide airflow smoothly along the aircraft's surface.

Reinforcement learning from human feedback (RLHF) and direct preference optimization (DPO) are important techniques to align large language models (LLM) with human preference. However, the quality of RLHF and DPO training is seriously compromised by \textit{\textbf{C}orrupted} preference, reward \textit{\textbf{O}veroptimization}, and bias towards \textit{\textbf{V}erbosity}. To our knowledge, most existing works tackle only one of these important issues, and the few other works require much computation to estimate multiple reward models and lack theoretical guarantee of generalization ability. In this work, we propose RLHF-\textbf{COV} and DPO-\textbf{COV} algorithms that can simultaneously mitigate these three issues, in both offline and online settings. This ability is theoretically demonstrated by obtaining length-regularized generalization error rates for our DPO-COV algorithms trained on corrupted data, which match the best-known rates for simpler cases with clean data and without length regularization. Moreover, our DPO-COV algorithm is simple to implement without reward estimation, and is proved to be equivalent to our RLHF-COV algorithm, which directly implies the equivalence between the vanilla RLHF and DPO algorithms. Experiments demonstrate the effectiveness of our DPO-COV algorithms under both offline and online settings.

This research report addresses the absence of an actionable definition for Loss of Control (LoC) in AI systems by developing a novel taxonomy and preparedness framework. Despite increasing policy and research attention, existing LoC definitions vary significantly in scope and timeline, hindering effective LoC assessment and mitigation. To address this issue, we draw from an extensive literature review and propose a graded LoC taxonomy, based on the metrics of severity and persistence, that distinguishes between Deviation, Bounded LoC, and Strict LoC. We model pathways toward a societal state of vulnerability in which sufficiently advanced AI systems have acquired or could acquire the means to cause Bounded or Strict LoC once a catalyst, either misalignment or pure malfunction, materializes. We argue that this state becomes increasingly likely over time, absent strategic intervention, and propose a strategy to avoid reaching a state of vulnerability. Rather than focusing solely on intervening on AI capabilities and propensities potentially relevant for LoC or on preventing potential catalysts, we introduce a complementary framework that emphasizes three extrinsic factors: Deployment context, Affordances, and Permissions (the DAP framework). Compared to work on intrinsic factors and catalysts, this framework has the unfair advantage of being actionable today. Finally, we put forward a plan to maintain preparedness and prevent the occurrence of LoC outcomes should a state of societal vulnerability be reached, focusing on governance measures (threat modeling, deployment policies, emergency response) and technical controls (pre-deployment testing, control measures, monitoring) that could maintain a condition of perennial suspension.

Machine learning-based closure models for LES have shown promise in capturing complex turbulence dynamics but often suffer from instabilities and physical inconsistencies. In this work, we develop a novel skew-symmetric neural architecture as closure model that enforces stability while preserving key physical conservation laws. Our approach leverages a discretization that ensures mass, momentum, and energy conservation, along with a face-averaging filter to maintain mass conservation in coarse-grained velocity fields. We compare our model against several conventional data-driven closures (including unconstrained convolutional neural networks), and the physics-based Smagorinsky model. Performance is evaluated on decaying turbulence and Kolmogorov flow for multiple coarse-graining factors. In these test cases we observe that unconstrained machine learning models suffer from numerical instabilities. In contrast, our skew-symmetric model remains stable across all tests, though at the cost of increased dissipation. Despite this trade-off, we demonstrate that our model still outperforms the Smagorinsky model in unseen scenarios. These findings highlight the potential of structure-preserving machine learning closures for reliable long-time LES.

Despite progress in machine learning-based aerodynamic surrogates, the scarcity of large, field-resolved datasets limits progress on accurate pointwise prediction and reproducible inverse design for aircraft. We introduce BlendedNet++, a large-scale aerodynamic dataset and benchmark focused on blended wing body (BWB) aircraft. The dataset contains over 12,000 unique geometries, each simulated at a single flight condition, yielding 12,490 aerodynamic results for steady RANS CFD. For every case, we provide (i) integrated force/moment coefficients CL, CD, CM and (ii) dense surface fields of pressure and skin friction coefficients Cp and (Cfx, Cfy, Cfz). Using this dataset, we standardize a forward-surrogate benchmark to predict pointwise fields across six model families: GraphSAGE, GraphUNet, PointNet, a coordinate Transformer (Transolver-style), a FiLMNet (coordinate MLP with feature-wise modulation), and a Graph Neural Operator Transformer (GNOT). Finally, we present an inverse design task of achieving a specified lift-to-drag ratio under fixed flight conditions, implemented via a conditional diffusion model. To assess performance, we benchmark this approach against gradient-based optimization on the same surrogate and a diffusion-optimization hybrid that first samples with the conditional diffusion model and then further optimizes the designs. BlendedNet++ provides a unified forward and inverse protocol with multi-model baselines, enabling fair, reproducible comparison across architectures and optimization paradigms. We expect BlendedNet++ to catalyze reproducible research in field-level aerodynamics and inverse design; resources (dataset, splits, baselines, and scripts) will be released upon acceptance.