Researchers have developed a new framework that uses network analysis and machine learning to identify illicit financial transactions with improved accuracy and interpretability.
A new study explores how advanced artificial intelligence techniques can identify subtle anomalies in the Canadian stock market, potentially predicting extreme events before they occur.
A new forecasting model leverages transformer networks and multimodal data to anticipate structural responses and improve wind turbine health monitoring.
New research details a system for mathematically proving the safety and compliance of autonomous AI agents operating within financial markets.
![Trajectory-persistent adversarial attacks reveal that recurrent state space model (RSSM) architectures amplify initial perturbations-increasing by a factor of 2.26× in the deterministic GRU world model-before GRU contraction attenuates them, a phenomenon not observed in single-step baselines, and which is mitigated through adversarial fine-tuning-reducing amplification across all steps-resulting in a reward gap of only [latex]0.000892 \pm 0.000057[/latex] at a planning horizon of 30, and demonstrating a fundamental trade-off between model expressiveness and robustness to adversarial input.](https://arxiv.org/html/2604.01346v1/x1.png)
As artificial intelligence builds increasingly complex models of reality, new safety vulnerabilities emerge from the systems’ imagined environments.
A new agentic architecture is emerging that moves beyond traditional algorithmic investing, offering a pathway to fully autonomous portfolio management.
A new approach leverages large language models and real-time news analysis to forecast disruptions and improve supply chain resilience.
![The SIGN framework demonstrates successful equation discovery across diverse networked dynamical systems - including Kuramoto phase-oscillator networks, susceptible-infected-susceptible (SIS) epidemic models, Michaelis-Menten regulatory networks, FitzHugh-Nagumo neuron models, and Hindmarsh-Rose neuron models - consistently inferring coefficient values with low error rates across varying network sizes and topologies, from synthetic scale-free networks ([latex]10^3[/latex] and [latex]10^5[/latex] nodes) to large empirical datasets like GitHub, Catster, and a human brain network.](https://arxiv.org/html/2604.00599v1/x2.png)
Researchers have developed a novel framework that combines the power of data and physics-based modeling to forecast the long-term evolution of massive, interconnected systems.
![The study dissects the confidence scaling of large language models-specifically GPT-5, DeepSeek-V3.2-Exp, and Mistral-Medium-2508-across three distinct tasks, revealing disparities in their ability to align reported confidence levels with task accuracy, as evidenced by metrics like [latex]d'\relax[/latex] and [latex]\text{Mrati}\relax[/latex], and further refined by the exclusion of outlier data points-approximately 0.1% for Mistral-Medium-2508 in task B-to ensure a robust assessment of confidence calibration across a trial count of [latex]2 \times \Gamma_{3}0^{\relax}[/latex] for task A and [latex]\Gamma_{3}0^{\relax}[/latex] for tasks B and C.](https://arxiv.org/html/2603.29693v1/x1.png)
New research explores whether large language models possess the capacity for metacognition – the ability to assess their own confidence and uncertainty.
As increasingly complex AI systems begin to collaborate, unforeseen and potentially harmful behaviors can arise from the interactions of individually rational agents.