I will discuss the following statement, a definable version of the (p,q) theorem of Jiří Matoušek from combinatorics, conjectured by Chernikov and Simon:

Suppose that T is NIP and that phi(x,b) does not fork over a model M. Then there is some formula psi(y) in tp(b/M) such that the partial type {phi(x,b’) : psi(b’)} is consistent.

I will describe the interaction between a single soliton and a gas of solitons, providing for the first time a mathematical justification for the kinetic theory as posited by Zakharov in the 1970s.  Then I will explain how to use random matrix theory to introduce randomness into a large collection of solitons.

Matthew Buckland 
Branching Interval Partition Diffusions

We construct an interval-partition-valued diffusion from a collection of excursions sampled from the excursion measure of a real-valued diffusion, and we use a spectrally positive Lévy process to order both these excursions and their start times. At any point in time, the interval partition generated is the concatenation of intervals where each excursion alive at that point contributes an interval of size given by its value. Previous work by Forman, Pal, Rizzolo and Winkel considers self-similar interval partition diffusions – and the key aim of this work is to generalise these results by dropping the self-similarity condition. The interval partition can be interpreted as an ordered collection of individuals (intervals) alive that have varying characteristics and generate new intervals during their finite lifetimes, and hence can be viewed as a class of Crump-Mode-Jagers-type processes.

Ofir Gorodetsky
Smooth and rough numbers

We all know and love prime numbers, but what about smooth and rough numbers?
We'll define y-smooth numbers -- numbers whose prime factors are all less than y. We'll explain their application in cryptography, specifically to factorization of integers.
We'll shed light on their density, which is modelled using a peculiar differential equation. This equation appears naturally in probability theory.
We'll also explain the dual notion to smooth numbers, that of y-rough numbers: numbers whose prime factors are all bigger than y, and in some sense generalize primes.
We'll explain their importance in sieve theory. Like smooth numbers, their density has interesting properties and will be surveyed.

Your supervisor is the person you will interact with on a scientific level most of all during your studies here. As a result, it is vital that you establish a good working relationship. But how should you do this? In this session we discuss tips and tricks for getting the most out of your supervisions to maximize your success as a researcher. Note that this session will have no faculty in the audience in order to allow people to speak openly about their experiences. 

What should we be thinking about when we're making a diagram for a paper? How do we help it to express the right things? Or make it engaging? What kind of colour palette is appropriate? What software should we use? And how do we make this process as painless as possible? Join Joshua Bull and Christoph Dorn for a lively Fridays@4 session on illustrating mathematics, as they share tips, tricks, and their own personal experiences in bringing mathematics to life via illustrations.

Generative AI has seen tremendous successes recently, most notably the chatbot ChatGPT and the DALLE2 software creating realistic images and artwork from text descriptions. Underlying these and other generative AI systems are usually neural networks trained to produce text, images, audio, or video from text inputs. The aim of this talk is to develop an understanding of the fundamental capabilities of generative neural networks. Specifically and mathematically speaking, we consider the realization of high-dimensional random vectors from one-dimensional random variables through deep neural networks. The resulting random vectors follow prescribed conditional probability distributions, where the conditioning represents the text input of the generative system and its output can be text, images, audio, or video. It is shown that every d-dimensional probability distribution can be generated through deep ReLU networks out of a 1-dimensional uniform input distribution. What is more, this is possible without incurring a cost—in terms of approximation error as measured in Wasserstein-distance—relative to generating the d-dimensional target distribution from d independent random variables. This is enabled by a space-filling approach which realizes a Wasserstein-optimal transport map and elicits the importance of network depth in driving the Wasserstein distance between the target distribution and its neural network approximation to zero. Finally, we show that the number of bits needed to encode the corresponding generative networks equals the fundamental limit for encoding probability distributions (by any method) as dictated by quantization theory according to Graf and Luschgy. This result also characterizes the minimum amount of information that needs to be extracted from training data so as to be able to generate a desired output at a prescribed accuracy and establishes that generative ReLU networks can attain this minimum.

This is joint work with D. Perekrestenko and L. Eberhard


 

Ilia Smilga
Margulis spacetimes and crooked planes

We are interested in the following problem: which groups can act 
properly on R^n by affine transformations, or in other terms, can occur 
as a symmetry group of a "regular affine tiling"? If we additionally 
require that they preserve a Euclidean metric (i.e. act by affine 
isometries), then these groups are well-known: they all contain a 
finite-index abelian subgroup. If we remove this requirement, a 
surprising result due to Margulis is that the free group can act 
properly on R^3. I shall explain how to construct such an action.

Charles Parker
Unexpected Behavior in Finite Elements for Linear Elasticity
One of the first problems that finite elements were designed to approximate is the small deformations of a linear elastic body; i.e. the 2D/3D version of Hooke's law for springs from elementary physics. However, for nearly incompressible materials, such as rubber, certain finite elements seemingly lose their approximation power. After briefly reviewing the equations of linear elasticity and the basics of finite element methods, we will spend most of the time looking at a few examples that highlight this unexpected behavior. We conclude with a theoretical result that (mostly) explains these findings.

Academic research can be challenging and can bring with it difficulties in maintaining good mental health. This session will be led by Rebecca Reed, Mental Health First Aid (MHFA) Instructor, Meditation & Yoga Teacher and Personal Development Coach and owner of wellbeing company Siendo. Rebecca will talk about how we can maintain good mental fitness, recognizing good practices to ensure we avoid mental-health difficulties before they begin. We have deliberately set this session to be at the beginning of the academic year in this spirit. We will also talk about maintaining good mental health specifically in the academic community.   

The data revolution has changed the landscape of computational mathematics and has increased the demand for new numerical linear algebra tools to handle the vast amount of data. One crucial task is data compression to capture the inherent structure of data efficiently. Tensor-based approaches have gained significant traction in this setting by exploiting multilinear relationships in multiway data. In this talk, we will describe a matrix-mimetic tensor algebra that offers provably optimal compressed representations of high-dimensional data. We will compare this tensor-algebraic approach to other popular tensor decomposition techniques and show that our approach offers both theoretical and numerical advantages.

Abstract: 

Welcome (back) to Fridays@4! To start the new academic year in this session we’ll introduce what Fridays@4 is for our new students and colleagues. This session will be a chance to meet current students and ECRs from across Maths and Stats who will share their hints and tips on conducting successful research in Oxford. There will be lots of time for questions, discussions and generally meeting more people across the two departments – everyone is welcome!