Rating Sustainability: Can AI Build Trust?
A new framework leverages the power of artificial intelligence and human expertise to assess the reliability of corporate sustainability ratings.
Science
Learning from the Past: AI Improves with Memory-Enhanced Reinforcement
A new framework empowers artificial intelligence to learn more efficiently by retrospectively analyzing past experiences and assigning credit where it’s due.
Mapping Dementia: A New Network Approach to Early Diagnosis
![PaIRWaL encodes random walks on similarity networks derived from gyral folding to produce invariant sequences, subsequently aggregated to facilitate graph-level classification-a process fundamentally reliant on the topological properties of the underlying data and yielding representations insensitive to node permutations [latex] \mathbb{R}^n [/latex].](https://arxiv.org/html/2602.17557v1/x1.png)
Researchers are leveraging the unique folding patterns of the brain to build more accurate diagnostic tools for Alzheimer’s and Lewy body dementia.
When Algorithms Get It Wrong: Navigating Imperfect Models
New research reveals how Bayesian algorithms like Thompson Sampling behave when faced with inaccurate models, offering insights into their robustness and potential pitfalls.
Riding the Chaos: Accurately Predicting System Responses to Complex Forces
![The system’s state space trajectories, captured at intervals of approximately 10 seconds, demonstrate the predictive capacity of a Gaussian State Space model (red) and an autoencoder-LSTM network (green) against the backdrop of a periodically forced system (cyan) and its true response (black) as projected onto different phase space coordinates-specifically, [latex]x_{10}, \dot{x}_{10}, x_{11}[/latex] and [latex]x_1, \dot{x}_1, x_{20}[/latex]-revealing the models’ ability to approximate system behavior across varying states.](https://arxiv.org/html/2602.16848v1/x16.png)
A new computational approach offers a faster and more precise method for determining the stable states of mechanical systems under unpredictable external influences.
Taming Diffusion Models: Pruning for Efficiency
A new pruning strategy targets unstable attention patterns in diffusion models to significantly improve computational efficiency without sacrificing performance.
Bridging Physics and Learning: A New Approach to System Modeling
A novel framework combines established physical models with data-driven techniques to achieve more accurate and efficient identification of complex nonlinear systems.
Predicting People: A New Model for Understanding Behavior
Researchers have developed a novel approach to forecasting individual actions by integrating psychological traits with the power of large language models.
Catching the Shift: A New Way to Spot Tipping Points in Complex Systems
A novel data-driven framework offers a robust method for detecting subtle structural changes that signal critical transitions in high-dimensional dynamical systems.
Building a Digital X-Ray: Visualizing Structural Damage in 3D
A new approach utilizes advanced 3D reconstruction techniques to create detailed digital twins of civil infrastructure, enabling precise damage assessment and long-term monitoring.