Adaptive agents through active inference: The main fields of research that are used to model and realise adaptive agents are optimal control, reinforcement learning and active inference. Active inference is a probabilistic description of adaptive agents that is relatively less known to mathematicians, as it originated from neuroscience in the last decade. This talk presents the mathematical underpinnings of active inference, starting from fundamental considerations about agents that maintain their structural integrity in the face of environmental perturbations. Through this, we derive a probability distribution over actions, that describes decision-making under uncertainty in adaptive agents . Interestingly, this distribution has an interesting information geometric structure, combining, for instance, drives for exploration and exploitation, which may yield a principled answer to the exploration-exploitation trade-off. Preserving this geometric structure enables to realise adaptive agents in practice. We illustrate their behaviour with simulation examples and empirical comparisons with reinforcement learning.

Scaling Limits of Random Graphs: The scaling limit of directed random graphs remains relatively unexplored compared to their undirected counterparts. In contrast, many real-world networks, such as links on the world wide web, financial transactions and “follows” on Twitter, are inherently directed. Previous work by Goldschmidt and Stephenson established the scaling limit for the strongly connected components (SCCs) of the Erdős -- Rényi model in the critical window when appropriately rescaled. In this talk, we present a result showing the SCCs of another class of critical random directed graphs will converge when rescaled to the same limit. Central to the proof is an exploration of the directed graph and subsequent encodings of the exploration as real valued random processes. We aim to present this exploration algorithm and other key components of the proof.

From Mathematics to Data Science and Back: We give an overview of the interaction between rough path theory and data science at the current time.
 

I will report on recent progress concerning eigenvalues of Schrödinger operators with complex potentials. We are interested in the magnitude and distribution of eigenvalues, and we seek bounds that only depend on an L^p norm of the potential.

These questions are well understood for real potentials, but completely new phenomena arise for complex potentials. I will explain how techniques from harmonic analysis, particularly those related to Fourier restriction theory, can be used to prove upper and lower bounds. We will also discuss some open problems. The talk is based on recent joint work with Sabine Bögli (Durham).

I will discuss 3d N=4 supersymmetric gauge theories compactified on an elliptic curve, and how this set-up physically realises recent mathematical results on the equivariant elliptic cohomology of symplectic resolutions. In particular, I will describe the Berry connection for supersymmetric ground states, and in doing so connect the elliptic cohomology of the Higgs branch with spectral data of doubly periodic monopoles. I will show that boundary conditions, via a consideration of boundary ’t Hooft anomalies, naturally represent elliptic cohomology classes. Finally, if I have time, I will discuss mirror symmetry/symplectic duality in our framework, and physically recover concepts in elliptic cohomology such as the mother function, and the elliptic stable envelopes of Aganagic-Okounkov.

This talk will be based on https://arxiv.org/abs/2109.10907 with Mathew Bullimore.

A well-known conjecture of Ivanov states that mapping class groups of surfaces with genus at least 3 virtually do not surject onto the integers. Putman and Wieland reformulated this conjecture in terms of higher Prym representations of finite-index subgroups of mapping class groups. We show that the Putman-Wieland conjecture holds for geometrically uniform subgroups. Along the way we construct a cover S of the genus 2 surface such that the lifts of simple closed curves do not generate the rational homology of S. This is joint work with Markovic.

Group theoretic Dehn filling, motivated by Dehn filling in the theory of 3- manifolds, is a process of constructing quotients of a given group. This technique is usually applied to groups with certain negative curvature feature, for example word-hyperbolic groups, to construct exotic and useful examples of groups. In this talk, I will start by recalling the notion of word-hyperbolic groups, and then show that how group theoretic Dehn filling can be used to answer the Burnside Problem and questions about mapping class groups of surfaces.

A big mapping class group is the mapping class group (MCG) of a surface of infinite type. Although several aspects of big MCGs remain mysterious, their geometric definition allows some simple, interesting arguments. In this talk, we will use big MCGs as an excuse to survey some (more or less) classical results in geometric group theory: we will present a quick introduction to infinite type surfaces, highlight differences between standard and large MCGs, and use Higman’s embedding theorem to deduce that there exists a big MCG that contains every finitely presented group as a subgroup.

We are confronted with signals defined on the nodes of a graph in many applications.  Think for instance of a sensor network measuring temperature; or a social network, in which each person (node) has an opinion about a specific issue.  Graph signal processing (GSP) tries to device appropriate tools to process such data by generalizing classical methods from signal processing of time-series and images -- such as smoothing, filtering and interpolation -- to signals defined on graphs.  Typically, this involves leveraging the structure of the graph as encoded in the spectral properties of the graph Laplacian.

In other applications such as traffic network analysis, however, the signals of interest are naturally defined on the edges of a graph, rather than on the nodes. After a very brief recap of the central ideas of GSP, we examine why the standard tools from GSP may not be suitable for the analysis of such edge signals.  More specifically, we discuss how the underlying notion of a 'smooth signal' inherited from (the typically considered variants of) the graph Laplacian are not suitable when dealing with edge signals that encode flows.  To overcome this limitation we devise signal processing tools based on the Hodge-Laplacian and the associated discrete Hodge Theory for simplicial (and cellular) complexes.  We discuss applications of these ideas for signal smoothing, semi-supervised and active learning for edge-flows on discrete (or discretized) spaces.

TBA