Untangling City Life: How Shared Causes Reveal Urban Events
A new framework leverages the underlying connections between different incidents to more accurately reconstruct what’s happening in a city based on resident reports.
Science
The Coordination Cost of Distributed Control
As complex engineering systems grow, distributing control via multi-agent reinforcement learning offers scalability, but new research reveals this can come at the expense of optimal performance.
Predicting Chemotherapy Success with the Power of Clinical Notes
New research demonstrates how analyzing doctor’s notes with advanced AI can forecast patient response to chemotherapy, potentially personalizing cancer treatment.
Unfolding the Future: A New Approach to Long-Term Traffic Prediction
Researchers have developed VisiFold, a novel framework that leverages graph-based techniques to dramatically improve the accuracy and efficiency of forecasting traffic patterns over extended periods.
Reading Minds, Shaping Markets
![The equilibrium mean control [latex]\bar{D}^{1}(t)[/latex] is demonstrably affected by signal precision [latex]p[/latex], exhibiting a heightened incentive for belief manipulation when precision is low due to sluggish opponent posteriors, and converging towards a perfect-information benchmark as [latex]p[/latex] approaches infinity.](https://arxiv.org/html/2603.12140v1/x3.png)
New research reveals how strategic players can both predict and influence the expectations of others, with significant implications for economic efficiency.
Orchestrating the Flow: Secure AI in Distributed Networks
This review explores a novel architecture for managing decentralized AI workloads across dynamic, multi-domain environments.
Intelligent Networks: Automating Operations with AI Agents
A new approach leverages the power of artificial intelligence to streamline the complex tasks of managing and maintaining modern optical networks.
When Rare Events Skew Reality: Fixing Human Labeling in AI
New research reveals how cognitive biases impact the accuracy of human-annotated data for rare-event AI, and proposes effective strategies to counteract them.
Who Do You Trust Online?
This review explores the methods used to quantify trust within the complex networks of online social platforms.
Beyond Confidence: Taming Classifier Uncertainty
New research demonstrates how improving the accuracy of prediction probabilities can reduce the number of equally valid, yet different, model predictions.