Seismic Sleuthing: AI Learns to Tell Earthquakes from Explosions
A new deep learning architecture combines seismological expertise with artificial intelligence to dramatically improve the accurate classification of seismic events.
Science
Global Flood Forecasts: A Machine Learning Leap
A new data-driven model, AIFL, delivers competitive streamflow predictions by leveraging the power of pre-trained machine learning and readily available weather data.
Listening for Trouble: AI-Powered Fault Detection in 3D Printers
A new approach uses sound and machine learning to identify mechanical failures in 3D printers as they happen.
Modeling the UK’s Carbon Budget: An Economic Reality Check
A new agent-based model assesses the macroeconomic and distributional consequences of the UK’s Seventh Carbon Budget, offering crucial insights for policymakers.
Streaming Graphs, Summarized: Crane Captures the Signal
![Crane’s architecture establishes a unified framework for large language model (LLM) serving, integrating components for request handling, model management, and efficient resource allocation to optimize throughput and minimize latency, as formalized in [latex] T = \frac{1}{N} \sum_{i=1}^{N} t_i [/latex], where <i>T</i> represents average latency and <i>t<sub>i</sub></i> denotes the processing time for each request.](https://arxiv.org/html/2602.15360v1/x1.png)
A new neural sketch, Crane, efficiently distills the essential information from continuously evolving graph data, offering a significant leap forward in stream summarization.
Agents That Think Ahead: Building More Reliable Web Automation
A new framework combines predictive modeling and reasoning to create web agents capable of anticipating errors and improving task completion in complex online environments.
Beyond the Baseline: Stress-Testing Anomaly Detection for Real-World IoT
New research reveals that current anomaly detection models struggle with the complexities of real-world IoT deployments, necessitating a more rigorous and realistic benchmarking approach.
Predicting Heart Failure with the Power of AI
A new language model analyzes electrocardiograms to forecast cardiac events, offering a promising step toward proactive heart health management.
Orchestrating Trust in Federated AI
A new framework leverages composable contracts to guarantee performance and accountability across distributed AI services.
AI-Powered Wireless: Fortifying Networks Against Emerging Threats
The convergence of agentic AI and advanced wireless networks promises enhanced security and privacy, but also introduces new vulnerabilities that demand careful consideration.