The memory function description of quantum operator dynamics involves a carefully chosen split into 'fast' and 'slow' modes. An approximate model for the fast modes can then be used to solve for Green's functions $G(z)$ of the slow modes. One common approach is to form a Krylov operator basis $\{O_m \}_{m\ge 0}$, and then approximate the dynamics restricted to the 'fast space' $\{O_m \}_{m\ge n}$ for $n\gg 1$. Denoting the fast space Green's function by $G_n (z)$, in this work we prove that $G_n (z)$ exhibits universality in the $n \to \infty$ limit, approaching different universal scaling forms in different regions of the complex $z$-plane. This universality of $G_n (z)$ is precisely analogous to the universality of eigenvalue correlations in random matrix theory (RMT). At finite $z$, we show that $G_n (z)$ approaches the Wigner semicircle law, the same as the average bulk resolvent for the Gaussian Unitary Ensemble. When $G(z)$ is the Green's function of certain hydrodynamical variables, we show that at low frequencies $G_n (z)$ is instead governed by the Bessel universality class from RMT. As an application we give a new numerical method, the spectral bootstrap, for approximating spectral functions from Lanczos coefficients. Our proof is complex analytic in nature, involving a Riemann-Hilbert problem which we solve using a steepest-descent-type method, controlled in the $n\to\infty$ limit. This proof technique assumes some regularity conditions on the spectral function, and we comment on their possible connections to the eigenstate thermalization hypothesis. We are led via steepest-descent to a related Coulomb gas ensemble, and we discuss how a recent conjecture--the 'Operator Growth Hypothesis'--is linked to a confinement transition in this Coulomb gas. We then explain the implications of this criticality for the computational resources required to estimate transport coefficients.

Background: Adolescents are particularly vulnerable to mental disorders, with over 75% of cases manifesting before the age of 25. Research indicates that only 18 to 34% of young people experiencing high levels of depression or anxiety symptoms seek support. Digital tools leveraging smartphones offer scalable and early intervention opportunities. Objective: Using a novel machine learning framework, this study evaluated the feasibility of integrating active and passive smartphone data to predict mental disorders in non-clinical adolescents. Specifically, we investigated the utility of the Mindcraft app in predicting risks for internalising and externalising disorders, eating disorders, insomnia and suicidal ideation. Methods: Participants (N=103; mean age 16.1 years) were recruited from three London schools. Participants completed the Strengths and Difficulties Questionnaire, the Eating Disorders-15 Questionnaire, Sleep Condition Indicator Questionnaire and indicated the presence/absence of suicidal ideation. They used the Mindcraft app for 14 days, contributing active data via self-reports and passive data from smartphone sensors. A contrastive pretraining phase was applied to enhance user-specific feature stability, followed by supervised fine-tuning. The model evaluation employed leave-one-subject-out cross-validation using balanced accuracy as the primary metric. Results: The integration of active and passive data achieved superior performance compared to individual data sources, with mean balanced accuracies of 0.71 for SDQ-High risk, 0.67 for insomnia, 0.77 for suicidal ideation and 0.70 for eating disorders. The contrastive learning framework stabilised daily behavioural representations, enhancing predictive robustness. This study demonstrates the potential of integrating active and passive smartphone data with advanced machine-learning techniques for predicting mental health risks.

We demonstrate that the machine learning of density functionals allows one to determine simultaneously the equilibrium chemical potential across simulation datasets of inhomogeneous classical fluids. Minimization of a loss function based on an Euler-Lagrange equation yields both the universal one-body direct correlation functional, which is represented locally by a neural network, as well as the system-specific unknown chemical potential values. The method can serve as an efficient alternative to conventional computational techniques of measuring the chemical potential. It also facilitates using canonical data from Brownian dynamics, molecular dynamics, or Monte Carlo simulations as a basis for constructing neural density functionals, which are fit for accurate multiscale prediction of soft matter systems in equilibrium.

The turnpike property refers to the phenomenon that in many optimal control problems, the solutions for different initial conditions and varying horizons approach a neighborhood of a specific steady state, then stay in this neighborhood for the major part of the time horizon, until they may finally depart. While early observations of the phenomenon can be traced back to works of Ramsey and von Neumann on problems in economics in 1928 and 1938, the turnpike property received continuous interest in economics since the 1960s and recent interest in systems and control. The present chapter provides an introductory overview of discrete-time and continuous-time results in finite and infinite-dimensions. We comment on dissipativity-based approaches and infinite-horizon results, which enable the exploitation of turnpike properties for the numerical solution of problems with long and infinite horizons. After drawing upon numerical examples, the chapter concludes with an outlook on time-varying, discounted, and open problems.

Inhomogeneities in the velocity field of a moving fluid are dampened by the inherent viscous behaviour of the system. Both bulk and shear effects, related to the divergence and the curl of the velocity field, are relevant. On molecular time scales, beyond the Navier-Stokes description, memory plays an important role. We demonstrate here on the basis of molecular and overdamped Brownian dynamics many-body simulations that analogous viscous effects act on the acceleration field. This acceleration viscous behaviour is associated with the divergence and the curl of the acceleration field and it can be quantitatively described using simple exponentially decaying memory kernels. The simultaneous use of velocity and acceleration fields enables the description of fast dynamics on molecular scales.

We consider a linear regression model with complex-valued response and predictors from a compact and connected Lie group. The regression model is formulated in terms of eigenfunctions of the Laplace-Beltrami operator on the Lie group. We show that the normalized Haar measure is an approximate optimal design with respect to all Kiefer's $\Phi_p$-criteria. Inspired by the concept of $t$-designs in the field of algebraic combinatorics, we then consider so-called $\lambda$-designs in order to construct exact $\Phi_p$-optimal designs for fixed sample sizes in the considered regression problem. In particular, we explicitly construct $\Phi_p$-optimal designs for regression models with predictors in the Lie groups $\mathrm{SU}(2)$ and $\mathrm{SO}(3)$, the groups of $2\times 2$ unitary matrices and $3\times 3$ orthogonal matrices with determinant equal to $1$, respectively. We also discuss the advantages of the derived theoretical results in a concrete biological application.

This thesis investigates the central role of homomorphism problems (structure-preserving maps) in two complementary domains: database querying over finite, graph-shaped data, and constraint solving over (potentially infinite) structures. Building on the well-known equivalence between conjunctive query evaluation and homomorphism existence, the first part focuses on conjunctive regular path queries, a standard extension of conjunctive queries that incorporates regular-path predicates. We study the fundamental problem of query minimization under two measures: the number of atoms (constraints) and the tree-width of the query graph. In both cases, we prove the problem to be decidable, and provide efficient algorithms for a large fragment of queries used in practice. The second part of the thesis lifts homomorphism problems to automatic structures, which are infinite structures describable by finite automata. We highlight a dichotomy, between homomorphism problems over automatic structures that are decidable in non-deterministic logarithmic space, and those that are undecidable (proving to be the more common case). In contrast to this prevalence of undecidability, we then focus on the language-theoretic properties of these structures, and show, relying on a novel algebraic language theory, that for any well-behaved logic (a pseudovariety), whether an automatic structure can be described in this logic is decidable.

In AI-assisted decision-making, a central promise of having a human-in-the-loop is that they should be able to complement the AI system by overriding its wrong recommendations. In practice, however, we often see that humans cannot assess the correctness of AI recommendations and, as a result, adhere to wrong or override correct advice. Different ways of relying on AI recommendations have immediate, yet distinct, implications for decision quality. Unfortunately, reliance and decision quality are often inappropriately conflated in the current literature on AI-assisted decision-making. In this work, we disentangle and formalize the relationship between reliance and decision quality, and we characterize the conditions under which human-AI complementarity is achievable. To illustrate how reliance and decision quality relate to one another, we propose a visual framework and demonstrate its usefulness for interpreting empirical findings, including the effects of interventions like explanations. Overall, our research highlights the importance of distinguishing between reliance behavior and decision quality in AI-assisted decision-making.

This paper contributes in the first part to the correct understanding of the linear limit in the Schamel equation (S-equation) from the perspective of structure formation in collisionless plasmas. The corresponding modes near equilibrium turn out to be nonlinear modes of the underlying microscopic Vlasov-Poisson (VP) system for which particle trapping is responsible. A simple shift of the electrostatic potential to a new pedestal leads to non-negativity and thus mitigates the positivity problem of the S-equation. The stability of a solitary electron hole (bright soliton), based on both the S-equation and an earlier transverse but limited VP instability analysis, exhibits marginal stability and linear perturbations in the form of the asymmetric shift eigenmode of a solvable Schrödinger problem. This finding of a dominant shift mode perturbation also seems to have been observed in a numerical PIC simulation, thus providing some confirmation for both theories.

This thesis investigates the central role of homomorphism problems (structure-preserving maps) in two complementary domains: database querying over finite, graph-shaped data, and constraint solving over (potentially infinite) structures. Building on the well-known equivalence between conjunctive query evaluation and homomorphism existence, the first part focuses on conjunctive regular path queries, a standard extension of conjunctive queries that incorporates regular-path predicates. We study the fundamental problem of query minimization under two measures: the number of atoms (constraints) and the tree-width of the query graph. In both cases, we prove the problem to be decidable, and provide efficient algorithms for a large fragment of queries used in practice. The second part of the thesis lifts homomorphism problems to automatic structures, which are infinite structures describable by finite automata. We highlight a dichotomy, between homomorphism problems over automatic structures that are decidable in non-deterministic logarithmic space, and those that are undecidable (proving to be the more common case). In contrast to this prevalence of undecidability, we then focus on the language-theoretic properties of these structures, and show, relying on a novel algebraic language theory, that for any well-behaved logic (a pseudovariety), whether an automatic structure can be described in this logic is decidable.

Based on results in finite geometry we prove the existence of MRD codes in (F_q)_(n,n) with minimum distance n which are essentially different from Gabidulin codes. The construction results from algebraic structures which are closely related to those of finite fields. Some of the results may be known to experts, but to our knowledge have never been pointed out explicitly in the literature.

For a curve of genus at least four which is either very general or very general hyperelliptic, we classify all ways in which a power of its Jacobian can be isogenous to a product of Jacobians of curves. As an application, we show that, for a very general principally polarized abelian variety of dimension at least four, or the intermediate Jacobian of a very general cubic threefold, no power is isogenous to a product of Jacobians of curves. This confirms some cases of the Coleman-Oort conjecture. We further deduce from our results some progress on the question whether the integral Hodge conjecture fails for such abelian varieties.

We prove three exact sum rules that relate the polarization of active Brownian particles to their one-body current: (i) The total polarization vanishes, provided that there is no net flux through the boundaries, (ii) at any planar wall the polarization is determined by the magnitude of the bulk current, and (iii) the total interface polarization between phase-separated fluid states is rigorously determined by the gas-liquid current difference. This result precludes the influence of the total interface polarization on active bulk coexistence and questions the proposed coupling of interface to bulk.

Optically confined colloidal particles, when placed in close proximity, form a dissipatively coupled system through hydrodynamic interactions. The role of such interactions influencing irreversibility and energy dissipation in out-of-equilibrium systems is often not well deciphered. Here, we demonstrate - through the estimation of the entropy production rate - that the nonequilibrium features of the system with such interactions vary depending on the nature of external driving, and importantly, on the level of coarse-graining. Crucially, we show that coarse-graining reverses the dependence of the measured entropy production rate on the strength of the hydrodynamic interactions. Furthermore, we clarify that such interactions do not violate energy balance at the level of individual trajectories, as was believed earlier. Our results highlight a previously unnoticed effect of coarse-graining in nonequilibrium systems, and have implications for the inference of entropy production in experimental contexts.

Fractal behavior and long-range dependence have been observed in an astonishing number of physical systems. Either phenomenon has been modeled by self-similar random functions, thereby implying a linear relationship between fractal dimension, a measure of roughness, and Hurst coefficient, a measure of long-memory dependence. This letter introduces simple stochastic models which allow for any combination of fractal dimension and Hurst exponent. We synthesize images from these models, with arbitrary fractal properties and power-law correlations, and propose a test for self-similarity.

If a two-level system coupled to a single-mode cavity is strongly driven by an external laser, instead of a continuous accumulation of photons in the cavity, oscillations in the mean photon number occur. These oscillations correspond to peaks of finite width running up and down in the photon number distribution, reminiscent of wave packets in linear chain models. A single wave packet is found if the cavity is resonant to the external laser. Here, we show that for finite detuning multiple packet structures can exist simultaneously, oscillating at different frequencies and amplitudes. We further study the influence of dissipative effects resulting in the formation of a stationary state, which depending on the parameters can be characterized by a bimodal photon number distribution. While we give analytical limits for the maximally achievable photon number in the absence of any dissipation, surprisingly, dephasing processes can push the photon occupations towards higher photon numbers.

The interplay of charge, spin and lattice degrees of freedom is studied for quasi-one-dimensional electron and spin systems coupled to quantum phonons. Special emphasis is put on the influence of the lattice dynamics on the Peierls transition. Using exact diagonalization techniques the ground-state and spectral properties of the Holstein model of spinless fermions and of a frustrated Heisenberg model with magneto-elastic coupling are analyzed on finite chains. In the non-adiabatic regime a (T=0) quantum phase transition from a gapless Luttinger-liquid/spin-fluid state to a gapped dimerized phase occurs at a nonzero critical value of the electron/spin-phonon interaction. To study the nature of the spin-Peierls transition at finite temperatures for the infinite system, an alternative Green's function approach is applied to the magnetostrictive XY model. With increasing phonon frequency the structure factor shows a remarkable crossover from soft-mode to central-peak behaviour. The results are discussed in relation to recent experiments on $\rm CuGeO_3$.

We describe recent progress in the statistical mechanical description of many-body systems via machine learning combined with concepts from density functional theory and many-body simulations. We argue that the neural functional theory by Samm\"uller et al. [Proc. Nat. Acad. Sci. 120, e2312484120 (2023)] gives a functional representation of direct correlations and of thermodynamics that allows for thorough quality control and consistency checking of the involved methods of artificial intelligence. Addressing a prototypical system we here present a pedagogical application to hard core particle in one spatial dimension, where Percus' exact solution for the free energy functional provides an unambiguous reference. A corresponding standalone numerical tutorial that demonstrates the neural functional concepts together with the underlying fundamentals of Monte Carlo simulations, classical density functional theory, machine learning, and differential programming is available online at https://github.com/sfalmo/NeuralDFT-Tutorial.

Particle resuspension refers to the physical process by which solid particles deposited on a surface are, first, detached and, then, entrained away by the action of a fluid flow. In this study, we explore the dynamics of large and heavy spherical particles forming a complex sediment bed which is exposed to a laminar shear flow. For that purpose, we rely on fine-scale simulations based on a fully-resolved flow field around individual particles whose motion is explicitly tracked. Using statistical tools, we characterize several features: (a) the overall bed dynamics (e.g. the average particle velocity as a function of the elevation), (b) the evolution of the top surface of the sediment bed (e.g. distribution of the surface elevation or of the surface slope) and (c) the dynamics of individual particles as they detach from or re-attach to the sediment bed (including the frequency of these events, and the velocity difference / surface angle for each event). These results show that particles detach more frequently around the peaks in the top surface of the sediment bed and that, once detached, they undergo short hops as particles quickly sediment towards the sediment bed. A simple model based on the surface characteristics (including its slope and elevation) is proposed to reproduce the detachment ratio.