I will present a parametric finite element formulation for structure-preserving numerical methods. The approach introduces two scalar Lagrange multipliers and evolution equations for surface energy and volume, ensuring that the resulting schemes maintain the underlying geometric and physical structures. To illustrate the method, I will discuss two applications: surface diffusion and two-phase Stokes flow. By combining piecewise linear finite elements in space with structure-preserving second-order time discretizations, we obtain fully discrete schemes of high temporal accuracy. Numerical experiments confirm that the proposed methods achieve the expected accuracy while preserving surface energy and volume.

Maximum likelihood estimation is a ubiquitous task in statistics and its applications. The task is: given some observations of a random variable, find the distribution(s) in your statistical model which best explains these observations. A modern perspective on this classical problem is to study the "likelihood geometry" of a statistical model. By focusing on models which have a polynomial parametrization, i.e., lie on an algebraic variety, this perspective brings in tools, algorithms and invariants from algebraic geometry and combinatorics.

In this talk, I will explain some of the key ideas in likelihood geometry and discuss its recent application to the study of likelihood asymptotics, i.e., understanding likelihood estimation for very large or very small observation counts. Agostini et al. showed that these asymptotics can be modeled and understood using tools from tropical geometry, and they used this to completely describe the asymptotics for linear models. In our work, we use the same approach to treat the class of log-linear models (also known as Gibbs distributions or maximum entropy models) and give a complete and combinatorial description of the likelihood asymptotics under some conditions.

This talk is based on joint work with Emma Boniface (UC Berkeley) and Serkan Hoşten (San Francisco SU), available at: https://epubs.siam.org/doi/full/10.1137/24M1656839

An inexact framework for high-order adaptive regularization methods is presented, in which approximations may be used for the pth-order tensor, based on lower-order derivatives. Between each recalculation of the pth-order derivative approximation, a high-order secant equation can be used to update the pth-order tensor as proposed in (Welzel 2022) or the approximation can be kept constant in a lazy manner. When refreshing the pth-order tensor approximation after m steps, an exact evaluation of the tensor or a finite difference approximation can be used with an explicit discretization stepsize. For all the newly adaptive regularization variants, we retrieve standard complexity bound to reach a second-order stationary point.  Discussions on the number of oracle calls for each introduced variant are also provided. When p = 2, we obtain a second-order method that uses quasi-Newton approximations with optimal number of iterations bound. 

In continuum kinetic models of quasineutral plasmas, binary collisions between particles are represented by the bilinear Fokker-Planck collision operator. In full-F kinetic models, which solve for the entire particle probability distribution function, it is important to correctly capture this operator, which pushes the system towards thermodynamic equilibrium. We show a multi-species, conservative, finite element implementation of this operator, using the continuum Galerkin representation, in the Julia programming language. A Jacobian-free-Newton-Krylov solver is used to implement a backward-Euler time advance. We present several example problems that demonstrate the performance of the implementation, and we speculate on future applications.

In the numerical simulation of incompressible flows and elastic materials, it is often desirable to design discretisation schemes that preserve key structural properties of the underlying physical model. In particular, the conservation of angular momentum plays a critical role in accurately capturing rotational effects, and is closely tied to the symmetry of the stress tensor. Classical formulations such as the Stokes equations or linear elasticity can exhibit significant discrepancies when this symmetry is weakly enforced or violated at the discrete level.

This work focuses on mixed finite element methods that impose the symmetry of the stress tensor strongly, thereby ensuring exact conservation of angular momentum in the absence of body torques and couple stresses. We systematically study the effect of this constraint in both incompressible Stokes flow and linear elasticity, including anisotropic settings inspired by liquid crystal polymer networks. Through a series of benchmark problems—ranging from rigid body motions to transversely isotropic materials—we demonstrate the advantages of angular-momentum-preserving discretisations, and contrast their performance with classical elements.

Our findings reveal that strong symmetry enforcement not only leads to more robust a priori error estimates and pressure-independent velocity approximations, but also more reliable physical predictions in scenarios where angular momentum conservation is critical.

These insights advocate for the broader adoption of structure-preserving methods in computational continuum mechanics, especially in applications sensitive to rotational invariants.

Transfer learning is a machine learning technique that leverages knowledge acquired in one domain to improve learning in another, related task. It is a foundational method underlying the success of large language models (LLMs) such as GPT and BERT, which were initially trained for specific tasks. In this talk, I will demonstrate how reinforcement learning (RL), particularly continuous time RL, can benefit from incorporating transfer learning techniques, especially with respect to convergence analysis. I will also show how this analysis naturally yields a simple corollary concerning the stability of score-based generative diffusion models.

Based on joint work with Zijiu Lyu of UC Berkeley.