Past

This is joint work with Paul Bourgade and Benjamin McKenna (Courant Institute, NYU).
The elastic manifold is a paradigmatic representative of the class of disordered elastic systems. These models describe random surfaces with rugged shapes resulting from a competition between random spatial impurities (preferring disordered configurations), on the one hand, and elastic self-interactions (preferring ordered configurations), on the other. The elastic manifold model is interesting because it displays a depinning phase transition and has a long history as a testing ground for new approaches in statistical physics of disordered media, for example for fixed dimension by Fisher (1986) using functional renormalization group methods, and in the high-dimensional limit by Mézard and Parisi (1992) using the replica method.
We study the topology of the energy landscape of this model in the Mézard-Parisi setting, and compute the (annealed) topological complexity both of total critical points and of local minima. Our main result confirms the recent formulas by Fyodorov and Le Doussal (2020) and allows to identify the boundary between simple and glassy phases. The core argument relies on the analysis of the asymptotic behavior of large random determinants in the exponential scale.

In this talk I will discuss a striking duality, T-duality, we discovered between two seemingly unrelated objects: the hypersimplex and the m=2 amplituhedron. We draw novel connections between them and prove many new properties. We exploit T-duality to relate their triangulations and generalised triangles (maximal cells in a triangulation). We subdivide the amplituhedron into chambers as the hypersimplex can be subdivided into simplices - both enumerated by Eulerian numbers. Along the way, we prove several conjectures on the amplituhedron and find novel cluster-algebraic structures, e.g. a generalisation of cluster adjacency.

This is based on the joint work with Lauren Williams and Melissa Sherman-Bennett https://arxiv.org/abs/2104.08254.

This is joint work with Paul Bourgade and Benjamin McKenna (Courant Institute, NYU).

The elastic manifold is a paradigmatic representative of the class of disordered elastic systems. These models describe random surfaces with rugged shapes resulting from a competition between random spatial impurities (preferring disordered configurations), on the one hand, and elastic self-interactions (preferring ordered configurations), on the other. The elastic manifold model is interesting because it displays a depinning phase transition and has a long history as a testing ground for new approaches in statistical physics of disordered media, for example for fixed dimension by Fisher (1986) using functional renormalization group methods, and in the high-dimensional limit by Mézard and Parisi (1992) using the replica method. 

We study the topology of the energy landscape of this model in the Mézard-Parisi setting, and compute the (annealed) topological complexity both of total critical points and of local minima. Our main result confirms the recent formulas by Fyodorov and Le Doussal (2020) and allows to identify the boundary between simple and glassy phases. The core argument relies on the analysis of the asymptotic behavior of large random determinants in the exponential scale.

Mixed-precision algorithms combine low- and high-precision computations in order to benefit from the performance gains of reduced-precision while retaining good accuracy. In this talk we focus on explicit stabilised Runge-Kutta (ESRK) methods for parabolic PDEs as they are especially amenable to a mixed-precision treatment. However, some of the concepts we present can be extended more generally to Runge-Kutta (RK) methods in general.

Consider the problem $y' = f(t,y)$ and let $u$ be the roundoff unit of the low-precision used. Standard mixed-precision schemes perform all evaluations of $f$ in reduced-precision to improve efficiency. We show that while this approach has many benefits, it harms the convergence order of the method leading to a limiting accuracy of $O(u)$.

In this talk we present a more accurate alternative: a scheme, which we call $q$-order-preserving, that is unaffected by this limiting behaviour. The idea is simple: by using $q$ high-precision evaluations of $f$ we can hope to retain a limiting convergence order of $O(\Delta t^{q})$. However, the practical design of these order-preserving schemes is less straight-forward.

We specifically focus on ESRK schemes as these are low-order schemes that employ a much larger number of stages than dictated by their convergence order so as to maximise stability. As such, these methods require most of the computational effort to be spent for stability rather than for accuracy purposes. We present new $s$-stage order $1$ and $2$ RK-Chebyshev and RK-Legendre methods that are provably full-order preserving. These methods are essentially as cheap as their fully low-precision equivalent and they are as accurate and (almost) as stable as their high-precision counterpart.

--

A link for this talk will be sent to our mailing list a day or two in advance.  If you are not on the list and wish to be sent a link, please contact @email.

Consider an $n$ by $n$ random matrix $A$ with i.i.d entries. In this talk, we discuss some results on the magnitude of the smallest singular value of $A$, and, in particular, the problem of estimating the singularity probability when the entries of $A$ are discrete.

Riche—Williamson recently proved that the characters of tilting modules for GL_h are given by non-singular p-Kazhdan—Lusztig polynomials providing p>h.  This is equivalent to calculating the decomposition numbers for symmetric groups labelled by partitions with at most h columns.  We discuss how this result can be generalised to all cyclotomic quiver Hecke algebras via a new and explicit isomorphism between (truncations of) quiver Hecke algebras and Elias–Williamson’s diagrammatic endomorphism algebras of Bott–Samelson bimodules. 

This allows us to give an elementary and explicit proof of the main theorem of Riche–Williamson’s recent monograph and extend their categorical equivalence to all cyclotomic quiver Hecke algebras, thus solving Libedinsky–Plaza’s categorical blob conjecture.  Furthermore, it allows us to classify and construct the homogeneous simple modules of quiver Hecke algebras via BGG resolutions.   
 
This is joint work with A. Cox, A. Hazi, D.Michailidis, E. Norton, and J. Simental.  
 

Standard worst-case rounding error bounds of most numerical linear algebra algorithms grow linearly with the problem size and the machine precision. These bounds suggest that numerical algorithms could be inaccurate at large scale and/or at low precisions, but fortunately they are pessimistic. We will review recent advances in probabilistic rounding error analyses, which have attracted renewed interest due to the emergence of low precisions on modern hardware as well as the rise of stochastic rounding.

--

A link for this talk will be sent to our mailing list a day or two in advance.  If you are not on the list and wish to be sent a link, please contact @email.

Online social networks such asTwitter, Facebook, Instagram and TikTokserve as mediafor the spread of information between their users.We areinterested in developing models forthis information diffusion to gain a greater understanding of its drivers. Some models forthe spread ofonlinebehaviour and informationassume that the information behaves similarly to a virus, where infection is equally likely after each exposure, these dynamics are known as a simple contagion. In a simple contagion, the exposures are independent of each other.However,online adoption of some behaviour and content has been empirically observed to be more likely after multiple exposures from their network neighbours, the exposures are not independent of each other, we refer to this as a complex contagion.Analytically tractable descriptions of complex contagions havebeendeveloped for continuous-time dynamics. These extend mean-field and pair approximation methods to account for clustering in the network topologies; however, no such analogous treatments for discrete-time cascade processes exist using branching processes. We describe a novel definition of complex contagion adoption dynamics and show how to construct multi-type branching processeswhichaccount for clustering on networks. We achieve this by tracking the evolution of a cascade via different classes of clique motifs whichaccount for the different numbers of active, inactive and removed nodes. This description allows for extensive MonteCarlo simulations (which are faster than network-based simulations), accurate analytical calculation of cascade sizes, determination of critical behaviour and other quantities of interest

Consider an n by n random matrix A with i.i.d entries. In this talk, we discuss some results on the magnitude of the smallest singular value of A, and, in particular, the problem of estimating the singularity probability when the entries of A are discrete.

Neural controlled differential equations (Neural CDEs) are a continuous-time extension of recurrent neural networks (RNNs). They are considered SOTA for modelling functions on irregular time series, outperforming other ODE benchmarks (ODE-RNN, GRU-ODE-Bayes) in offline prediction tasks. However, current implementations are not suitable to be used in online prediction tasks, severely restricting the domains of applicability of this powerful modeling framework. We identify such limitations with previous implementations and show how said limitations may be addressed, most notably to allow for online predictions. We benchmark our online Neural CDE model on three continuous monitoring tasks from the MIMIC-IV ICU database, demonstrating improved performance on two of the three tasks against state-of-the-art (SOTA) non-ODE benchmarks, and improved performance on all tasks against our ODE benchmark.

Joint work with Patrick Kidger, Lingyi Yang, and Terry Lyons.

I will present a theorem on the full finite codimension nonlinear asymptotic stability of the Schwarzschild family of black holes.  The proof employs a double null gauge, is expressed entirely in physical space, and utilises the analysis of Dafermos--Holzegel--Rodnianski on the linear stability of the Schwarzschild family.  This is joint work with M. Dafermos, G. Holzegel and I. Rodnianski.

 Given a solution of the Einstein equations, a fundamental question is whether one can extend the solution or whether the solution is maximal. If the solution is inextendible in a certain regularity class due to the geometry becoming singular, a further question is whether the strength of the singularity is such that it terminates classical time-evolution. The latter question, as will be explained in the talk, is intimately tied to the strong cosmic censorship conjecture in general relativity which states in the language of partial differential equations that global uniqueness holds generically for the initial value problem for the Einstein equations. I will then focus in the talk on recent results showing the locally Lipschitz inextendibility of FLRW models with particle horizons and spherically symmetric weak null singularities. The latter in particular apply to the spherically symmetric spacetimes constructed by Luk and Oh, improving their C^2-formulation of strong cosmic censorship to a locally Lipschitz formulation.

A set of positive integers is primitive (or 1-primitive) if no member divides another. Erdős proved in 1935 that the weighted sum $\sum 1/(n\log n)$ for n ranging over a primitive set A is universally bounded over all choices for A. In 1988 he asked if this universal bound is attained by the set of prime numbers. One source of difficulty in this conjecture is that $\sum n^{-\lambda}$ over a primitive set is maximized by the primes if and only if $\lambda$ is at least the critical exponent $\tau_1\approx1.14$.
A set is $k$-primitive if no member divides any product of up to $k$ other distinct members. In joint work with C. Pomerance and T.H. Chan, we study the critical exponent $\tau_k$ for which the primes are maximal among $k$-primitive sets. In particular we prove that $\tau_2<0.8$, which directly implies the Erdős conjecture for 2-primitive sets.

The d-dimensional bordism category Cob_d has as objects closed (d-1)-manifolds and as morphisms diffeomorphism classes of d-dimensional bordisms. For d=1 and d=2 this category is well understood because we have a complete list of all 1 or 2-manifolds with boundary. In this talk I will argue that the categories Cob_1 and Cob_2 nevertheless carry a lot of interesting structure. 

I will show that the classifying spaces B(Cob_1) and B(Cob_2) contain interesting moduli spaces coming from the combinatorics of how 1 or 2 manifolds can be glued along their boundary. In particular, I will introduce the notion of a "factorisation category" and explain how it relates to Connes' cyclic category for d=1 and to the moduli space of tropical curves for d=2. If time permits, I will sketch how this relates to the curve complex and moduli spaces of complex curves.

It is widely believed that consistent theories of quantum gravity satisfy two basic kinematic constraints: they are free from any global symmetry, and they contain a complete spectrum of gauge charges. For compact, abelian gauge groups, completeness follows from the absence of a 1-form global symmetry. However, this correspondence breaks down for more general gauge groups, where the breaking of the 1-form symmetry is insufficient to guarantee a complete spectrum. We show that the correspondence may be restored by broadening our notion of symmetry to include non-invertible topological operators, and prove that their absence is sufficient to guarantee a complete spectrum for any compact, possibly disconnected gauge group. In addition, we prove an analogous statement regarding the completeness of twist vortices: codimension-2 objects defined by a discrete holonomy around their worldvolume, such as cosmic strings in four dimensions. I will also discuss how this correspondence is modified in more general contexts, including e.g. Chern-Simons terms. 

We use machine learning to approximate Calabi-Yau and SU(3)-structure metrics, including for the first time complex structure moduli dependence. Our new methods furthermore improve existing numerical approximations in terms of accuracy and speed. Knowing these metrics has numerous applications, ranging from computations of crucial aspects of the effective field theory of string compactifications such as the canonical normalizations for Yukawa couplings, and the massive string spectrum. In the case of SU(3) structure, our machine learning approach allows us to engineer metrics with certain torsion properties. Our methods are demonstrated for Calabi-Yau and SU(3)-structure manifolds based on a one-parameter family of quintic hypersurfaces in ℙ4.

I briefly give an overview on the key ML frameworks involved in this analysis (neural networks, auto-differentiation). This talk is mainly based on 2012.04656.

Clemens Koppensteiner
Categorifying Heisenberg algebras

Categorification replaces set-theoretic structures with category-theoretic analogues. We discuss what this means and why it is useful. We then discuss recent work on categorifying Heisenberg algebras and their Fock space representations. In particular this gives a satisfying answer to an observation about equivariant K-theory made by Ian Grojnowski in 1996.

Aggregation-Diffusion Equations
David Gómez-Castro

The aim of this talk is to discuss an evolution problem modelling particles systems exhibiting aggregation and diffusion phenomena, and we will focus mostly on the so-called Aggregation-Diffusion Equation: ∂ρ ∂t = ∇ · (ρ ∇(U′ (ρ) + V + W ∗ ρ)) (ADE)

First, we will discuss the modelling. The famous case U′ (ρ) = log ρ and W = 0 is the famous Heat Equation. In the classical literature, the term U′(ρ) is typically deduced from Darcy’s law and models an internal energy of the system. We will show through particle systems how the term V models a confinement energy and W ∗ ρ an aggregation energy. The complete model covers many famous examples from different disciplines: Porous Media, Fokker-Plank, Keller-Segel and others. After this modelling, we discuss the mathematical treatment of (ADE). As in the case of the Heat Equation, the diffusion cases where W = V = 0 are typically studied in the Lebesgue and Sobolev spaces. However, as in the Keller-Segel problem, a Dirac measures may appear in finite time. We present the Wasserstein distance between measures, which is a natural framework for these equations, connecting with the theory of Optimal Transport. In fact, when U, V and W are convex, (ADE) can be studied as the gradient-flow of a free-energy functional (i.e. curves minimising this energy) in this Wasserstein distance, applying Calculus of Variations techniques. We will discuss the minimisation problem associated to F, with an interest to the existence of Dirac measures. Finally, we will present new results showing that indeed, in some cases besides Keller-Segel, states with a Delta can be achieved through solutions of the evolution problem

 In this work we systematically introduce relations to topological data analysis (TDA) in the categories of sets, simplicial complexes and vector spaces to characterize and study the general dynamical behaviors in a consistent way. The proposed framework not only offers new insights to the classical TDA methodologies, but also motivates new approaches to interesting applications of TDA in dynamical metric spaces, dynamical coverings, etc. The associated algorithm which produces barcode invariants, and relations in more general categories will also be discussed.

A well-behaved field theory living on a fixed background has a causality structure defined by the background metric. In holography, however, signals can travel through the bulk, and some bulk metrics would allow a signal to travel faster than the speed of light as seen on the boundary. These are called boundary causality violating metrics. Holographers usually work with a classical bulk metric, in which case they declare that boundary causality violating metrics are forbidden. However, in a full quantum gravity path integral, these metrics do contribute. The question is then: how to avoid causality violation in this context? In this talk I will give a prescription that achieves this.

One of the most successful methods in topological data analysis (TDA) is persistent homology, which associates a one-parameter family of spaces to a data set, and gives a summary — an invariant called "barcode" — of how topological features, such as the number of components, holes, or voids evolve across the parameter space. In many applications one might wish to associate a multiparameter family of spaces to a data set. There is no generalisation of the barcode to the multiparameter case, and finding algebraic invariants that are suitable for applications is one of the biggest challenges in TDA.

The use of persistent homology in applications is justified by the validity of certain stability results. At the core of such results is a notion of distance between the invariants that one associates to data sets. While such distances are well-understood in the one-parameter case, the study of distances for multiparameter persistence modules is more challenging, as they rely on a choice of suitable invariant.

In this talk I will first give a brief introduction to multiparameter persistent homology. I will then present a general framework to study stability questions in multiparameter persistence: I will discuss which properties we would like invariants to satisfy, present different ways to associate distances to such invariants, and finally illustrate how this framework can be used to derive new stability results. No prior knowledge on the subject is assumed.

The talk is based on joint work with Barbara Giunti, John Nolan and Lukas Waas. 

The graph C*-algebra construction associates a unital C*-algebra to any directed graph with finitely many vertices and countably many edges in a way which generalizes the fundamental construction by Cuntz and Krieger. We say that two such graphs are C*-equivalent when they define isomorphic C*-algebras, and give a description of this relation as the smallest equivalence relation generated by a number of "moves" on the graph that leave the C*-algebras unchanged. The talk is based on recent work with Arklint and Ruiz, but most of these moves have a long history that I intend to present in some detail.

Model-free machine learning is a tabula rasa method, estimating parametric functions purely from the data. In contrast, model-driven machine learning augments mathematical models with machine learning. For example, unknown terms in SDEs and PDEs can be represented by neural networks. We compare these two approaches, discuss their mathematical theory, and present several examples. In model-free machine learning, we use reinforcement learning to train order-execution models on limit order book data. Event-by-event simulation, based on the historical order book dataset, is used to train and evaluate limit order strategies. In model-driven machine learning, we develop SDEs and PDEs with neural network terms for options pricing as well as, in an application outside of finance, predictive modeling in physics. We are able to prove global convergence of the optimization algorithm for a class of linear elliptic PDEs with neural network terms.