Past

Numerically solving PDEs typically requires compressing infinite information into a finite system of algebraic equations. Pragmatically, we usually follow a recipe: “Assume solutions of form X; substitute into PDE Y; discard terms by rule Z.” In contrast, Lanczos’s pioneering “tau method” prescribes modifying the PDE to form an exact finite system. Crucially, any recipe-based method can be viewed as adding a small equation correction, enabling us to compare multiple schemes independently of the solver. 

This talk also addresses a paradox: PDEs often admit infinitely many solutions, but finite systems produce only a finite set. When we include a “small” correction, the missing solutions are effectively hidden. I will discuss how tau methods frame this perspective and outline proposals for systematically studying and optimising various residuals.

The notion of a superconformal structure on a supermanifold goes back some forty years. I will discuss some recent work that shows how these structures and their deformations govern supersymmetric and superconformal field theories in geometric fashion. A superconformal structure equips a supermanifold with a sheaf of dg commutative algebras; the tangent sheaf of this dg ringed space reproduces the Weyl multiplet of conformal supergravity (equivalently, the superconformal stress tensor multiplet), in any dimension and with any amount of supersymmetry. This construction is uniform under twists, and thus provides a classification of relations between superconformal theories, chiral algebras, higher Virasoro algebras, and more exotic examples.
 

Shape optimization is a rich field at the intersection of analysis, optimization, and engineering. It seeks to determine the optimal geometry of structures to minimize performance objectives, subject to physical constraints—often modeled by Partial Differential Equations (PDEs). Traditional approaches commonly assume that these constraints admit a unique solution for each candidate shape, implying a single-valued shape-to-solution map. However, many real-world structures exhibit multistability, where multiple stable configurations exist under identical physical conditions.

This research departs from the single-solution paradigm by investigating shape optimization in the presence of multiple solutions to the same PDE constraints. Focusing on a neo-Hookean hyperelastic model, we formulate an optimization problem aimed at controlling the energy jump between distinct solutions. Drawing on bifurcation theory, we develop a theoretical framework that interprets these solutions as continuous branches parameterized by shape variations. Building on this foundation, we implement a numerical optimization strategy and present numerical results that demonstrate the effectiveness of our approach.

Several physical models are available to understand the dynamics of fluid mixtures, including the so-called Baer-Nunziato models. The partial differential equations associated with these models look like those of Navier-Stokes, with the addition of new relaxation terms. One strategy to obtain these models is homogenisation: starting from a mesoscopic mixture, where two pure fluids satisfying the compressible Navier-Stokes equations share the space between them, a change of scale is performed to obtain a macroscopic mixture, where the two fluids can coexist at any point in space.

This problem concerns the study of the Navier-Stokes equations with strongly oscillating initial data. We'll start by explaining some results in this framework, in one dimension of space and on the torus, for barotropic fluids. We will then detail the various steps involved in demonstrating homogenisation. Finally, we'll explain how to adapt this reasoning to homogenisation for perfect gases, with and without heat conduction.

Predictions of complex flows remain a significant challenge for engineering systems. Computationally affordable predictions of turbulent flows generally require Reynolds-Averaged Navier–Stokes (RANS) simulations and Large-Eddy Simulation (LES), the predictive accuracy of which can be insufficient due to non-Boussinesq turbulence and/or unresolved multiphysics that preclude qualitative fidelity in certain regimes. For example, in turbulent combustion, flame–turbulence interactions can lead to inverse-cascade energy transfer, which violates the assumptions of many RANS and LES closures. We present an adjoint-based, solver-embedded data assimilation method to augment the RANS and LES equations using trusted data. This is accomplished using Python-native flow solvers that leverage differentiable programming techniques to construct the adjoint equations needed for optimization. We present applications to shock-tube ignition delay predictions, turbulent premixed jet flames, and shock-dominated nonequilibrium flows and discuss the potential of adjoint-based approaches for future machine learning applications.

I introduce modal group theory, where one investigates the class of all groups using embeddability as a modal operator. By employing HNN extensions, I demonstrate that the modal language of groups is more expressive than the first-order language of groups. Furthermore, I establish that the theory of true arithmetic, viewed as sets of Gödel numbers, is computably isomorphic to the modal theory of finitely presented groups. Finally, I resolve an open question posed by Sören Berger, Alexander Block, and Benedikt Löwe by proving that the propositional modal validities of groups constitute precisely the modal logic S4.2.

Recently, much attention has been paid to the intersection between coarse geometry and graph theory, giving rise to the fresh, exciting new field aptly known as ‘coarse graph theory’. One aspect of this area is the study of so-called ‘fat minors’, a large-scale analogue of the usual idea of a graph minor.

In this talk, I will introduce this area and motivate some interesting questions and conjectures. I will then sketch a proof that a finitely presented group is either virtually planar or contains arbitrarily ‘fat’ copies of every finite graph.

No prior knowledge or passion for graph theory will be assumed in this talk.

Given a smooth geometrically connected curve C over a perfect field k and a smooth commutative group scheme G defined over the function field K of C, one can consider isomorphism classes of G-torsors locally trivial at completions of K coming from closed points of C. They form a generalized Tate-Shafarevich group which specializes to the classical one in the case when k is finite. Recently, these groups have been studied over other base fields k as well, for instance p-adic or number fields. Surprisingly, finiteness can be proven in some cases but there are also quite a few open questions which I plan to discuss  in my talk.

Conformal welding, the process of glueing together Riemann surfaces along their boundaries, has recently played a prominent role in probability theory. In this talk, I will discuss two examples, namely the welding associated with random Jordan curves (SLE(k) loops) and particularly their limit as k tends to zero, and the welding of random trees (such as the CRT).

In this talk I will present two new results concerning differentiability of Lipschitz maps between Carnot groups. The former is a suitable adaptation of Pansu-Rademacher differentiability theorem to general Radon measures. More precisely we construct a suitable bundle associated to the measure along which Lipschitz maps are differentiable, very much in the spirit of the results of Alberti-Marchese. The latter is the converse of Pansu’s theorem. Namely, let G be a Carnot group and μ a Radon measure on G. Suppose further that every Lipschitz map between G and H, some other Carnot group, is Pansu differentiable μ-almost everywhere. We show that μ must be absolutely continuous with respect to the Haar measure of G. This is a joint work with Guido De Philippis, Andrea Marchese, Andrea Pinamonti and Filip Rindler.

This new sub-Riemannian result will be an excuse to present and discuss the techniques employed in Euclidean spaces to prove the converse of Rademacher's theorem.

The two-point Chowla conjecture predicts that $\sum_{x<n<2x} \lambda(n)\lambda(n+1) = o(x)$ as $x\to \infty$, where $\lambda$ is the Liouville function (a $\{\pm 1\}$-valued multiplicative function encoding the parity of the number of prime factors). While this remains an open problem, weaker versions of this conjecture are known. In this talk, we outline an approach initiated by Helfgott and Radziwill, which reformulates the problem in terms of bounding the eigenvalues of a certain matrix.

The Complexity Cluster Event Organisers Professor Gui-Qiang G. Chen, Professor Helen Byrne, and Professor Mohit Dalwadi, cordially invite you to attend a Complexity Cluster Research Workshop on Tuesday, 20th May 2025, in the Glen Callater Room, H B Allen Centre, Keble College.

Complexity Cluster Research Workshop
Venue: Glen Callater Room, H B Allen Centre, Keble College
Date: Tuesday, 20th May 2025
Organisers: 
Professor Helen Byrne
Professor Gui-Qiang G. Chen
Professor Mohit Dalwadi
 

Programme:
4.00pm ̶ 4.15pm: Coffee, Drinks & Refreshments
4:15pm ̶ 4:40pm: Professor Didier Bresch (CNRS and Universite Savoie Mont-Blanc, France): Mathematical Topics around Granular Media
4:45pm – 5:10pm: Dr. Keith Chambers (Mathematical Institute, University of Oxford): Structured Population Models to Explore Lipid-Driven Macrophage Heterogeneity in Early Atherosclerotic Plaques
5:15pm ̶ 5:40pm: Coffee, Drinks & Refreshments
5:40pm ̶ 6:05pm: Dr. Tara Trauthwein (Department of Statistics, University of Oxford): Approximation Results for Large Networks
6:05pm-6:30pm: Isaac Newell (OxPDE, Mathematical Institute, University of Oxford): The Gauss Equation for Isometric Embeddings of Regularity in W^1+2/3,3∩ C¹
6:35pm-7:00pm: Discussion

For abstracts please click the file here: Complexity-Cluster_Workshop_20250212_Final_0.pdf

Quantized universal enveloping algebras admit an intriguing class of (unbounded) Hilbert space representations obtained via their cluster structure. In these so-called positive representations the standard generators act by (essentially self-adjoint) positive operators. 

The aim of this talk is to discuss some analytical questions arising in this context, and in particular to what extent these representations can be understood using the theory of locally compact quantum groups in the sense of Kustermans and Vaes. I will focus on the simplest case in rank 1, where many of the key features (and difficulties) are already visible. (Based on work in progress with Kenny De Commer, Gus Schrader and Alexander Shapiro). 

The Complexity Cluster Event Organisers Professor Gui-Qiang G. Chen, Professor Helen Byrne, and Professor Mohit Dalwadi, cordially invite you to attend a Complexity Cluster Research Workshop on Tuesday, 20th May 2025, in the Glen Callater Room, H B Allen Centre, Keble College.

Complexity Cluster Research Workshop
Venue: Glen Callater Room, H B Allen Centre, Keble College
Date: Tuesday, 20th May 2025
Organisers: 
Professor Helen Byrne
Professor Gui-Qiang G. Chen
Professor Mohit Dalwadi
 

Programme:
4.00pm ̶ 4.15pm: Coffee, Drinks & Refreshments
4:15pm ̶ 4:40pm: Professor Didier Bresch (CNRS and Universite Savoie Mont-Blanc, France): Mathematical Topics around Granular Media
4:45pm – 5:10pm: Dr. Keith Chambers (Mathematical Institute, University of Oxford): Structured Population Models to Explore Lipid-Driven Macrophage Heterogeneity in Early Atherosclerotic Plaques
5:15pm ̶ 5:40pm: Coffee, Drinks & Refreshments
5:40pm ̶ 6:05pm: Dr. Tara Trauthwein (Department of Statistics, University of Oxford): Approximation Results for Large Networks
6:05pm-6:30pm: Isaac Newell (OxPDE, Mathematical Institute, University of Oxford): The Gauss Equation for Isometric Embeddings of Regularity in W^1+2/3,3∩ C¹
6:35pm-7:00pm: Discussion

For abstracts please click the file here: Complexity-Cluster_Workshop_20250212_Final_0.pdf

The Complexity Cluster Event Organisers Professor Gui-Qiang G. Chen, Professor Helen Byrne, and Professor Mohit Dalwadi, cordially invite you to attend a Complexity Cluster Research Workshop on Tuesday, 20th May 2025, in the Glen Callater Room, H B Allen Centre, Keble College.

Complexity Cluster Research Workshop
Venue: Glen Callater Room, H B Allen Centre, Keble College
Date: Tuesday, 20th May 2025
Organisers: 
Professor Helen Byrne
Professor Gui-Qiang G. Chen
Professor Mohit Dalwadi
 

Programme:
4.00pm ̶ 4.15pm: Coffee, Drinks & Refreshments
4:15pm ̶ 4:40pm: Professor Didier Bresch (CNRS and Universite Savoie Mont-Blanc, France): Mathematical Topics around Granular Media
4:45pm – 5:10pm: Dr. Keith Chambers (Mathematical Institute, University of Oxford): Structured Population Models to Explore Lipid-Driven Macrophage Heterogeneity in Early Atherosclerotic Plaques
5:15pm ̶ 5:40pm: Coffee, Drinks & Refreshments
5:40pm ̶ 6:05pm: Dr. Tara Trauthwein (Department of Statistics, University of Oxford): Approximation Results for Large Networks
6:05pm-6:30pm: Isaac Newell (OxPDE, Mathematical Institute, University of Oxford): The Gauss Equation for Isometric Embeddings of Regularity in W^1+2/3,3∩ C¹
6:35pm-7:00pm: Discussion

For abstracts please click the file here: Complexity-Cluster_Workshop_20250212_Final_0.pdf

A consequence of the works of Costello and Lurie is that the Hochschild chain complex of a Calabi-Yau category admits the structure of a framed E_2 algebra (the genus zero operations). I will describe a new algebraic point of view on these operations which admits generalizations to the setting of relative
Calabi-Yau structures, which do not seem to fit into the framework of TQFTs. In particular, we obtain a generalization of string topology to manifolds with boundary, as well as interesting operations on Hochschild homology of Fano varieties. This is joint work with Chris Brav.

Given an FP-infinity subgroup G of SL(2,C), we are interested in the asymptotic behavior of the cohomology of G with coefficients in an irreducible complex representation V of SL(2,C). We prove that, as the dimension of V grows, the dimensions of these cohomology groups approximate the L2-Betti numbers of G. We make no further assumptions on G, extending a previous result of W. Fu. This yields a new method to compute those Betti numbers for finitely generated hyperbolic 3-manifold groups. We will give a brief idea of the proof, whose main tool is a completion of the universal enveloping algebra of the p-adic Lie algebra sl(2, Zp).

Given an FP-infinity subgroup G of SL(2,C), we are interested in the asymptotic behavior of the cohomology of G with coefficients in an irreducible complex representation V of SL(2,C). We prove that, as the dimension of V grows, the dimensions of these cohomology groups approximate the L2-Betti numbers of G. We make no further assumptions on G, extending a previous result of W. Fu. This yields a new method to compute those Betti numbers for finitely generated hyperbolic 3-manifold groups. We will give a brief idea of the proof, whose main tool is a completion of the universal enveloping algebra of the p-adic Lie algebra sl(2, Zp).

Bestvina--Brady groups were first introduced by Bestvina and Brady for their interesting finiteness properties. In this talk, we discuss their Dehn functions, that are a notion of isoperimetric inequality for finitely presented groups and can be thought of as a "quantitative version" of finite presentability. A result of Dison shows that the Dehn function of a Bestvina--Brady group is always bounded above by a quartic polynomial.

Our main result is to compute the Dehn function for all finitely presented Bestvina--Brady groups. In particular, we show that the Dehn function of a Bestvina--Brady group grows as a polynomial of integer degree, and we present the combinatorial criteria on the graph that determine whether the Dehn functions of the associated Bestvina--Brady group is linear, quadratic, cubic, or quartic.

This is joint work with Chang and García-Mejía.

We consider quantum electrodynamics in 2+1 dimensions (QED3) with N matter fields and Chern-Simons level k. For small values of k and N, this theory describes various experimentally relevant systems in condensed matter, and is also conjectured to be part of a web of non-supersymmetric dualities. We compute the scaling dimensions of monopole operators in a large N and k expansion, which appears to be extremely accurate even down to the smallest values of N and k, and allows us to find dynamical evidence for these dualities and make predictions about the phase transitions. For instance, we combine these estimates with the conformal bootstrap to predict that the notorious Neel-VBS transition (QED3 with 2 scalars) is tricritical, which was recently confirmed by independent lattice simulations. Lastly, we propose a novel phase diagram for QED3 with 2 fermions, including duality with the O(4) Wilson-Fisher fixed point.

When dealing with PDEs arising in fluid mechanics, bounded-energy weaksolutions are, in many cases, the only type of solutions for which one can guarantee global existence without imposing any restrictions on the size of the initial data or forcing terms. Understanding how to construct such solutions is also crucial for designing stable numerical schemes.

In this talk, we will explain the strategy for contructing weak solutions for the Navier-Stokes system for viscous compressible flows, emphasizing the difficulties encountered in the case of non-isotropic viscous stress tensors. In particular, I will present some results obtained in collaboration with Didier Bresch and Maja Szlenk.

Given a mod $p$ Galois representation, one often wonders whether it arises by reducing a $p$-adic one, and whether these lifts are suitably 'well-behaved'. In this talk, we discuss how ideas from homotopy theory aid the study of Galois deformations, reviewing work of Galatius-Venkatesh.

Kreck’s modified surgery gives an approach to classify 2n-manifolds up to stable diffeomorphism, i.e., up to a connected sum with copies of $S^n \times  S^n$. In dimension 4, we use a combination of modified and classical surgery to compare the stable diffeomorphism classification with other stable equivalence relations. Most importantly, we consider homotopy equivalence up to connected sum with copies of $S^2 \times  S^2$. This talk is based on joint work with John Nicholson and Simona Veselá.

In this seminar, we explore the quantitative convergence of wide deep neural networks with Gaussian weights to Gaussian processes, establishing novel rates for their Gaussian approximation. We show that the Wasserstein distance between the network output and its Gaussian counterpart scales inversely with network width, with bounds apply for any finite input set under specific non-degeneracy conditions of the covariances. Additionally, we extend our analysis to the Bayesian framework, by studying exact posteriors for neural networks, when endowed with Gaussian priors and regular Likelihood functions, but we also provide recent advancements in quantitative approximation of trained networks via gradient descent in the NTK regime. Based on joint works with A. Basteri, and A. Agazzi and E. Mosig.