Brief

Researchers have developed a novel physics-informed generative framework to model yield curve dynamics, addressing the conflict between deep learning's flexibility and fixed-income modeling's theoretical constraints. The proposed two-stage architecture, featuring a Student-t Conditional Variational Autoencoder with Dynamic Level Injection (CVAEsT+LS) and a Neural Stochastic Differential Equation penalized by a No-Arbitrage PDE, significantly reduces forecasting errors. This approach demonstrates superior performance in predicting term structures across various macroeconomic regimes and currencies, outperforming traditional models like HJM. AI

Researchers have developed a new method to improve the efficiency of training neural likelihood surrogates for stochastic process models. By augmenting the standard loss function with exact score information and adaptive weighting, the approach significantly reduces the computational cost associated with parameter inference. This technique demonstrates improved surrogate quality and can achieve performance comparable to a tenfold increase in training data with only a marginal increase in training time. AI

Kevin Warsh has been confirmed as the new Federal Reserve chair in a 54-45 Senate vote, largely along party lines. He takes the helm amid resurgent inflation, economic uncertainty, and unprecedented challenges to the Fed's independence. Warsh, a former Fed official, has been critical of the institution's past policies and has promised significant changes, though he faces a divided rate-setting committee and pressure from President Trump regarding interest rates. AI

Alibaba's AI business has entered a commercialization phase, with annualized recurring revenue from its AI models and applications, including the Baichuan MaaS platform, projected to exceed 10 billion yuan in the June quarter and reach 30 billion yuan by year-end. This growth is driven by increasing demand from enterprise clients for model and application services, evidenced by a significant rise in token consumption. Concurrently, storage company Bowei Storage has refiled its application for an H-share listing in Hong Kong, though the offering remains subject to regulatory approvals. AI

Researchers have demonstrated a method to aggregate asymptotically optimal sequential tests into log-optimal e-processes. This work proves the converse of a previous finding, establishing that optimal sequential tests can indeed be constructed from these e-processes. The new approach utilizes a novel class of WAIT e-processes, which are weighted aggregates of indicators of stopping times. AI

Researchers have developed a new theoretical framework for neural operators, a type of AI model used to learn solutions for complex systems like partial differential equations. This work specifically addresses the approximation analysis for nonlinear reaction-diffusion systems, which are crucial for modeling pattern formation. The study establishes explicit error bounds and demonstrates that their proposed Laplacian eigenfunction-based architecture can significantly reduce the parameter complexity required for accurate predictions. AI

Public health officials are monitoring a hantavirus outbreak that originated on a cruise ship, with no current signs of a widespread pandemic. Despite a high mortality rate for this strain, its spread requires close contact, making a COVID-19-level crisis unlikely. Meanwhile, NVision Quantum Technologies has secured $55 million in new funding to advance its quantum-based metabolic imaging for real-time cancer treatment monitoring, valuing the company at over $250 million. AI

Researchers have introduced Random-Set Graph Neural Networks (RS-GNNs) to address uncertainty quantification in graph learning. This new framework models node-level epistemic uncertainty using a belief function formalism. Experiments on nine datasets, including autonomous driving benchmarks, show RS-GNNs offer improved uncertainty estimation capabilities. AI

Researchers have introduced Quantized Diffusion Schrödinger Bridges (QDSB), a novel method for learning generative models from unpaired data. QDSB addresses the computational challenges of traditional Schrödinger bridges by quantizing endpoint distributions and using cell-wise sampling to reconstruct the data plan. This approach significantly reduces training time while maintaining sample quality comparable to existing methods. AI

Researchers have developed a new spectral perspective to better understand tree ensemble algorithms like random forests and gradient boosting machines. This approach reveals that the decay rate of eigenvalues in the induced kernel operator dictates the statistical convergence for random forest regression. The findings also enable the creation of compressed tree ensembles, yielding significantly smaller models that retain competitive predictive accuracy, outperforming current methods for forest pruning and rule extraction. AI

Negotiations between Samsung Electronics and its largest labor union have collapsed, with an 18-day strike at its chip factories now imminent. The union is demanding changes to bonus calculations, including removing a cap and allocating a percentage of annual profit, while Samsung has only offered a one-time payment. The potential strike, projected to cost $700 million daily, has prompted an emergency meeting called by the South Korean Prime Minister due to significant economic implications, as semiconductors form a large portion of the country's exports. AI

Edge computing is becoming crucial for modern manufacturing, enabling real-time data analysis and split-second decision-making by moving processing power closer to machines. This shift is projected to drive over $380 billion in global spending by 2028, with a significant portion of data processed outside traditional data centers. AI further enhances edge capabilities by facilitating predictive maintenance, optimizing workflows, and improving energy efficiency, while also bolstering cybersecurity measures against rising industrial threats. AI

Two new research papers introduce novel generative approaches to improve pose estimation accuracy. The first, GenCape, uses a structure-aware variational autoencoder and graph transfer module to infer keypoint relationships from limited examples without predefined skeletons. The second paper focuses on 3D human pose estimation by employing controllable generative augmentation to synthesize diverse video data, systematically varying poses, backgrounds, and camera viewpoints to enhance domain generalization. AI

A Nevada electric company plans to divert power from 50,000 residents to supply data centers. This decision affects multiple towns, which have been told to find alternative power sources. The move highlights the growing demand for energy to support data center infrastructure, particularly for AI. AI

Isomorphic Labs, an AI-driven drug discovery company, and Anduril, a defense technology firm, have both recently secured significant funding rounds. Isomorphic raised $2.1 billion in Series B funding, while Anduril closed a $5 billion Series H round at a $61 billion valuation. Thrive Capital was a lead investor in both rounds, with Andreessen Horowitz also leading for Anduril. Both companies are leveraging advancements in AI to pursue ambitious goals, with Isomorphic aiming to cure diseases and Anduril developing autonomous weapons systems. AI

Fourier Intelligence has signed strategic cooperation agreements with Singapore's NHG Health and Japan's Nagoya University to advance embodied healthcare. These collaborations aim to accelerate the clinical translation of rehabilitation and robotics technologies, focusing on personalized rehabilitation and scalable robotic applications. Fourier Intelligence also showcased its brain-computer interface integrated with lower limb exoskeletons, proposing a "brain-computer embodied intelligent rehabilitation port" concept to enhance rehabilitation efficiency. AI

Researchers have developed a theoretical framework for sparse Bayesian Kolmogorov-Arnold Networks (KANs). Their work establishes statistical foundations for KANs, demonstrating that these networks can achieve near-minimax posterior contraction rates. The analysis shows that KANs can adapt to unknown function smoothness and avoid the curse of dimensionality by controlling approximation complexity through width and parameter sparsity, rather than depth. AI

Researchers have developed a new active inference framework for U-statistics, aiming to improve estimation efficiency when labeling data is expensive. This approach selectively queries informative labels within a fixed budget, building upon augmented inverse probability weighting U-statistics. The framework is also extended to U-statistic-based empirical risk minimization, showing significant gains in efficiency and maintaining target coverage in experiments. AI

SoftBank Group announced a surprising increase in its quarterly profits, largely attributed to the appreciated valuation of its investment in OpenAI. This financial boost highlights the significant impact of AI company valuations on investment firms' performance. AI

Tilde Research has introduced Aurora, a novel optimizer designed to train neural networks more effectively. Aurora addresses a critical issue in the popular Muon optimizer where a significant number of neurons become permanently inactive during training. The new optimizer, demonstrated with a 1.1B parameter pretraining experiment, achieves state-of-the-art performance on the modded-nanoGPT speedrun benchmark and has its code released publicly. AI

Researchers have introduced a new framework called post-ADC inference to address the challenges of statistical validity when data collected through active data collection (ADC) is reused for subsequent inferential tasks. This method accounts for biases introduced by both the data collection process and data-dependent target construction. The framework aims to provide valid p-values and confidence intervals, applicable to various ADC processes without strict assumptions on the underlying black-box function or surrogate models. AI

U.S. authorities are investigating multiple cases of advanced Nvidia GPUs and other semiconductor technology being illegally smuggled to China and Russia, circumventing export controls. These efforts involve sophisticated schemes, including the use of fake front companies and encrypted communications, to move restricted chips for purposes ranging from AI development to military applications. Despite significant penalties and enforcement actions against companies like Applied Materials and Cadence Design Systems, the illicit flow of this technology continues, posing challenges to national security. AI

Researchers have developed new online prediction algorithms designed to adapt their calibration error based on the degree of non-stationarity in the environment. These algorithms aim to perform optimally across a spectrum from stable, i.i.d. settings to highly adversarial ones. The proposed methods achieve adaptive calibration guarantees, matching optimal rates in stationary cases and recovering known bounds for adversarial regimes. AI

A Chinese court has awarded over £28,000 in compensation to a worker who was fired after his company replaced him with AI. The Hangzhou intermediate people’s court ruled that the tech company was wrong to dismiss the employee, whose surname is Zhou, when he refused a demotion and pay cut after AI was implemented. This ruling, along with a similar case in Beijing, suggests a shift in Chinese policy towards protecting workers' job security amidst rapid AI adoption and high youth unemployment. AI

Philippine development groups are facing significant challenges a year after the US Agency for International Development (USAID) drastically reduced its contracts and funding. These cuts, which affected projects supporting democracy, human rights, environmental advocacy, and journalism, were reportedly influenced by misinformation campaigns. The US State Department has eliminated over 90% of USAID contracts globally, leading to a substantial loss of critical support for organizations in the Philippines. AI

The Korea Exchange (KRX) has begun integrating AI technology into its capital market surveillance systems. This move follows KRX's acquisition of AI startup Fair Labs, aimed at accelerating its AI transformation and bolstering its data operations. The adoption of AI is expected to enhance the efficiency and effectiveness of market monitoring. AI

Researchers have developed a new method for non-asymptotic quantization of spherically symmetric distributions, addressing limitations of Zador's theorem in high dimensions. The proposed approach utilizes random quantizers uniformly distributed on a sphere, achieving exceptional performance with moderate sample sizes. This method allows for precise computation of expected distortion and efficient numerical determination of the optimal radius, with approximations derived from extreme-value theory for scenarios where sample size scales with dimension. AI

Researchers have developed a new algorithm for identifying $\varepsilon$-good actions in fixed-budget Monte Carlo Tree Search (MCTS). This algorithm is $\varepsilon$-agnostic, meaning it does not require the error tolerance $\varepsilon$ as an input but still provides instance-dependent error bounds. The misidentification probability decays exponentially with the budget, and the analysis offers new guarantees for specific MCTS methods while highlighting differences in hardness compared to standard K-armed bandits. AI

Researchers have developed a new dual representation for influence functions, which can efficiently estimate changes in model parameters and outputs. This method scales with dataset size rather than model size, offering an advantage for large models where traditional influence function evaluation is infeasible. However, the approach is currently limited to linearizable models and requires substantial matrix materialization. AI

Cisco reported strong fourth-quarter financial results, exceeding analyst expectations for both earnings and revenue, largely driven by significant demand for AI infrastructure. The company secured $5.3 billion in AI-related orders and raised its full-year AI revenue outlook to $4 billion. Despite these positive financial indicators and a 14% stock surge, Cisco also announced workforce reductions affecting nearly 4,000 employees, which CEO Chuck Robbins stated are necessary to optimize costs in light of AI opportunities. AI

Researchers have developed a new interpretable kernel for Gaussian Processes that can model rotational anisotropy in 3D spatial fields. This kernel explicitly parameterizes principal length-scales and orientation, offering a more intuitive approach than standard axis-aligned methods or generic SPD metrics. The method was tested on synthetic data and a material-density dataset, showing improved predictive performance and the ability to reveal complex anisotropy not captured by existing techniques. AI

Researchers have introduced Variational Predictive Resampling (VPR), a new method designed to improve the accuracy of Bayesian posterior sampling. VPR leverages variational inference's predictive capabilities within a resampling framework to better approximate the true posterior distribution. This approach aims to overcome the limitations of standard variational inference, which can sometimes produce overly concentrated approximations that miss important posterior dependencies. Experiments show VPR significantly enhances uncertainty quantification and recovers missed posterior dependencies, while remaining computationally efficient compared to traditional MCMC methods. AI

Organized opposition to AI data center expansion is coalescing into a more structured, networked approach. A new website, Data Center Opposition, tracks over 268 local groups across 37 states, aiming to facilitate coordination and information sharing among communities concerned about data center development. This growing resistance is increasingly influencing permitting, rezoning, and deployment timelines, with notable examples in Virginia, Missouri, and Georgia where community campaigns have led to project delays, moratoriums, and even outright cancellations. AI

The Netherlands has formally protested a proposed US law that would expand restrictions on chip equipment manufacturer ASML's sales to China. The Dutch government specifically objects to the extraterritorial reach of the proposed "Match Act," which would not only ban ASML from selling lower-end lithography machines but also prevent servicing existing Chinese customers. These new restrictions would build upon existing US export controls that already prohibit ASML from selling its most advanced extreme ultraviolet machines to China. AI

Researchers have developed a new framework for Probabilistic Partial Least Squares (PPLS) that addresses practical limitations in existing fitting pipelines. This framework combines noise pre-estimation, constrained likelihood optimization, and prediction calibration, offering an end-to-end solution. The method utilizes exact Stiefel-manifold optimization and noise-subspace estimation, achieving improved accuracy and calibrated uncertainty across various benchmarks, including multi-omics datasets. AI

This series of articles details the process of building and training machine learning models within an MLOps framework. The initial posts focus on setting up the development environment, including creating Python virtual environments and integrating with GitHub Actions for automated workflows. The content is aimed at guiding individuals through the practical steps of MLOps. AI

Researchers have published a paper detailing uniform scaling limits in transformers trained with the AdamW optimizer. The study models hidden-state dynamics as an interacting particle system, demonstrating convergence to a forward-backward system of ODEs. This convergence rate is dependent on the transformer's depth and number of heads, with specific mathematical bounds derived that are independent of token count and embedding dimension. AI

A massive data center, projected to be twice the size of Manhattan, is planned for Utah. This facility will be powered by liquefied natural gas (LNG), a decision that has drawn criticism regarding its environmental impact and emissions. The project is moving forward despite concerns about its energy source and scale. AI

Xunfei's Doubao LLM is reportedly receiving enhanced capabilities, though specific details remain undisclosed. Separately, Scenovation Technology has secured nearly $100 million in Series C funding, led by Suzhou Industrial Park Investment Group, to advance its automotive and embodied AI chip development. Additionally, a report from Scotiabank predicts a global copper deficit of 350,000 tons by 2027, driven by robust demand and supply-side challenges. AI

As of April 30th, China has seen a 50% year-over-year increase in the number of registered private securities products, with over 5,400 products filed. This surge is largely driven by index-enhanced and multi-asset strategies, which have grown by over 80% and 66% respectively. Notably, a significant portion of these new filings, over 30%, comes from top-tier private fund managers, indicating a concentration of capital towards leading firms. AI

Researchers have introduced WorldLens, a new benchmark designed to evaluate the realism and behavioral fidelity of driving world models. Current models often excel in either visual realism or physical consistency but not both, creating a gap in how their performance is assessed. WorldLens addresses this by measuring aspects like pixel quality, 4D geometry, closed-loop driving, and human perceptual alignment across 24 dimensions. Evaluations using WorldLens revealed that no single model performs optimally across all criteria, highlighting the need for more comprehensive assessment tools. AI

Microsoft Research has unveiled GridSFM, a compact foundation model designed to optimize power grid efficiency. This model can predict optimal AC power flow in milliseconds, aiding operators in managing grid congestion, stability, and overall system health for cost savings. Separately, Andrew Ng refutes the notion of an imminent "jobpocalypse" due to AI, asserting that AI will transform rather than replace jobs, necessitating adaptation over excessive fear. AI

The upcoming meeting between U.S. President Donald Trump and Chinese President Xi Jinping in Beijing is expected to address critical technology issues, including the intense rivalry in artificial intelligence and the trade of advanced AI chips. Both nations are considering dialogues on AI safety, while simultaneously competing in AI development and facing accusations of illicit distillation techniques. The discussions will also cover supply chain security, electric vehicle trade, and the global market for rare earth minerals, with significant implications for companies like Nvidia and Chinese AI labs such as DeepSeek. AI

NVision, a startup specializing in cancer imaging, has secured $38 million in funding led by Abbott, with an additional $17 million venture loan. The company utilizes quantum technology to enhance MRI scans, allowing for real-time monitoring of a tumor's metabolic response to treatment. This innovation aims to provide faster feedback on therapy effectiveness, potentially changing treatment timelines from months to days. NVision plans to expand its metabolic imaging system to 20 centers globally and is preparing for human clinical studies in 2027. AI

Researchers have published a paper detailing concentration phenomena in mean-field transformers, specifically analyzing their behavior at low temperatures during inference. The study uses a mean-field continuity equation to model token evolution and demonstrates that token distributions rapidly concentrate under a projection map induced by the transformer's matrices. This concentration remains metastable for moderate times, with the Wasserstein distance scaling in relation to temperature and inference time. AI

Researchers have developed a novel approach to solve multi-agent path finding (MAPF) problems by reformulating them as a specific type of multi-marginal optimal transport (MMOT) problem. This method leverages a Markovian structure to reduce the computational complexity of MMOT to a polynomial-sized linear program. For large-scale applications, the approach is further adapted using Schrödinger bridges, which provide an iterative, Sinkhorn-type solution that significantly reduces complexity while maintaining near-optimal results. AI

快手计划将其AI部门可灵（Kling）分拆出来，并寻求20亿美元的融资。此举表明了公司对AI业务的重视以及对独立发展的期望。此前，快手已在AI领域进行了大量投入，但面临预算快速消耗的挑战。 AI

Researchers have developed a new neural exponential tilting framework for variational inference in Lévy-driven stochastic differential equations. This method addresses the intractability of Bayesian inference for processes with heavy tails and discontinuities, which are crucial for modeling extreme events in fields like finance and AI safety. The framework uses neural networks to reweight the Lévy measure, preserving jump structures while remaining computationally efficient and enabling more reliable posterior inference than Gaussian-based methods. AI

Researchers have developed DECO-MWE, a new linguistic resource for analyzing sentiment in Korean text, specifically focusing on multiword expressions (MWEs). This resource utilizes the Local Grammar Graph (LGG) methodology, formalizing MWEs as a Finite-State Transducer. The DECO-MWE lexicon categorizes MWEs into four types, including standard polarity, domain-dependent polarity, named entity, and feature MWEs, achieving an f-measure of 0.806 in test corpora. The methodology and lexicon are intended for broad application in feature-based sentiment analysis. AI

Researchers have analyzed the regularization effects of data augmentation on supervised regression methods, particularly in scenarios where the number of covariates scales with the number of samples. The study provides a precise characterization of test error, using mean squared error, based on population quantities of the true data and statistics of the augmentation process. These findings apply to models with misspecified feature maps and architectures where only the final layer is trained, with the rest of the network being fixed or randomly initialized. AI