The field of dynamical systems modeling and control is witnessing significant developments, with a focus on improving the accuracy, efficiency, and robustness of models and control strategies. Researchers are exploring innovative approaches to address long-standing challenges, such as the need for explicit knowledge of system parameters, costly retraining, and limited generalization capability. Meta-learning and structure-preserving methods are being investigated to enable scalable and generalizable learning across parametric families of dynamical systems. Additionally, data-driven approaches, such as nested operator inference and symplectic neural networks, are being developed to learn reduced-order models that preserve the underlying physical constraints and symmetries of the systems. These advances have the potential to impact a broad range of applications, from physics and engineering to climate modeling and control. Noteworthy papers in this area include:

• Meta-learning Structure-Preserving Dynamics, which introduces a modulation-based meta-learning framework for scalable and generalizable learning of dynamical systems.
• Reduced-order modeling of Hamiltonian dynamics based on symplectic neural networks, which presents a novel data-driven symplectic induced-order modeling framework for high-dimensional Hamiltonian systems.
• Universal Learning of Nonlinear Dynamics, which describes an algorithm for learning marginally stable unknown nonlinear dynamical systems using spectral filtering and online convex optimization.