arXiv subjects

Markus Mueller

Publications and source records attributed to Markus Mueller.

At least 19 recordsLinked to original sources

arXiv2026

Fractal Confinement and Magnetic Self-sabotage of Current Flow: Electrons Near the Metal-insulator Crossover in a 2D $\delta$-layer

We study low-$T$ transport and itinerant magnetism in 2D $\delta$-doped semiconductors. Large-scale Hartree-Fock calculations with random hoppings and repulsion $U_0$ reveal two transport regimes near a critical density. The conductivity $\sigma(T)$ initially increases as $T$ drops below $U_0 / 10$. However, at $T \sim U_0/100$, a tiny density of magnetic moments forms, driving an insulating downturn. Fractal wavefunctions locally amplify magnetic interactions, initially concentrating currents into branched filaments, but ultimately choking them off via dynamic SU(2) symmetry breaking.

cond-mat.dis-nn↗

arXiv2026

Automating Visual Recognition of Leprosy in Wild Chimpanzees

Leprosy (Mycobacterium leprae) has been confirmed in wild western chimpanzees (Pan troglodytes verus) in West Africa, presenting as clear and progressive visual symptoms. Manual review of camera-trap footage at landscape scale is infeasible, motivating the need for automated screening. We present the first deep learning pipeline for wildlife leprosy detection and contribute the PanLep300 dataset of 125,670 annotated bounding-box crops across 953 tracks from 303 camera-trap videos with ecologically-motivated splits that withhold whole individuals and camera installations. We benchmark spatial (2D), temporally aggregated (2.5D), and video-based (3D) classification approaches to investigate which approach is best suited to automated leprosy detection in wild apes. We find that simple aggregation of crop-level predictions consistently matches or outperforms both learned temporal models and end-to-end video architectures -- consistent with leprosy's static cutaneous presentation. We further find that performance is suppressed when tracklets contain frames of partially visible individuals -- as commonly occurs at the start and end of a track -- and demonstrate that this can be addressed through targeted construction and aggregation strategies.

cs.CV↗

arXiv2026

Cascaded Flow Matching for Heterogeneous Tabular Data with Mixed-Type Features

Advances in generative modeling have recently been adapted to tabular data containing discrete and continuous features. However, generating mixed-type features that combine discrete states with an otherwise continuous distribution in a single feature remains challenging. We advance the state-of-the-art in diffusion models for tabular data with a cascaded approach. We first generate a low-resolution version of a tabular data row, that is, the collection of the purely categorical features and a coarse categorical representation of numerical features. Next, this information is leveraged in the high-resolution flow matching model via a novel guided conditional probability path and data-dependent coupling. The low-resolution representation of numerical features explicitly accounts for discrete outcomes, such as missing or inflated values, and therewith enables a more faithful generation of mixed-type features. We formally prove that this cascade tightens the transport cost bound. The results indicate that our model generates significantly more realistic samples and captures distributional details more accurately, for example, the detection score improves by 51.9\%. Code is available at https://github.com/muellermarkus/tabcascade.

cs.LG↗

arXiv2023

Continuous Diffusion for Mixed-Type Tabular Data

Score-based generative models, commonly referred to as diffusion models, have proven to be successful at generating text and image data. However, their adaptation to mixed-type tabular data remains underexplored. In this work, we propose CDTD, a Continuous Diffusion model for mixed-type Tabular Data. CDTD is based on a novel combination of score matching and score interpolation to enforce a unified continuous noise distribution for both continuous and categorical features. We explicitly acknowledge the necessity of homogenizing distinct data types by relying on model-specific loss calibration and initialization schemes. To further address the high heterogeneity in mixed-type tabular data, we introduce adaptive feature- or type-specific noise schedules. These ensure balanced generative performance across features and optimize the allocation of model capacity across features and diffusion time. Our experimental results show that CDTD consistently outperforms state-of-the-art benchmark models, captures feature correlations exceptionally well, and that heterogeneity in the noise schedule design boosts sample quality. Replication code is available at https://github.com/muellermarkus/cdtd.

cs.LG↗

arXiv2022

Photoinduced sliding transition into a hidden phase in van der Waals materials

We propose a generic scenario for metastability and excitation-induced switching in layered materials. Focusing on a minimal bilayer stack, where each layer consists of a honeycomb lattice with A and B sublattices, we map out the energy landscape with respect to the relative sliding of the layers. The sliding affects the interlayer hopping, which induces a splitting between bonding and anti-bonding bands. When this splitting is large, the AA and AB stacking configurations correspond to the global and secondary minima, respectively, and these configurations are separated by a barrier against layer-sliding. While chemical doping only flattens this barrier, strong \emph{photodoping} from bonding to antibonding bands can transiently destabilize the global minimum and induce a sliding motion toward the AB stacked configuration, thereby switching from an equilibrium insulator to a nearly gapless metastable phase. This hopping-driven effect is enhanced by local repulsive interactions, which increase the gap and facilitate the inter-layer sliding.

cond-mat.str-el↗

arXiv2018

Collision-dominated conductance in clean 2D metals

We study the temperature-dependent corrections to the conductance due to electron-electron (e-e) interactions in clean two-dimensional conductors, such as lightly doped graphene or other Dirac matter. We use semiclassical Boltzmann kinetic theory to solve the problem of collision-dominated transport between reflection-free contacts. Time-reversal symmetry and the kinematic constraints of scattering in two dimensions (2D) ensure that inversion-odd and inversion-even distortions of the quasiparticle distribution relax with parametrically different rates at low temperature. This entails the surprising result that at lowest temperatures the conductance of very long samples tends to the noninteracting, ballistic conductance, despite the relaxation of the quasiparticle distribution to a drifting equilibrium. The relative correction to the conductance depends on the ratio of relaxation rates of even and odd modes and scales as delta G/G_{ballistic}~(T/E_F)[Log(E_F/T)]^{1/2}, in stark contrast to the behavior in other dimensionalities. This holds generally in 2D systems with simply connected and convex but otherwise arbitrary Fermi surfaces, as long as e-e scattering processes are dominant and umklapp scattering is negligible. These results are especially relevant to the bulk of wide and long suspended high-mobility graphene sheets.

cond-mat.mes-hall↗

arXiv2018

Linguistic unit discovery from multi-modal inputs in unwritten languages: Summary of the "Speaking Rosetta" JSALT 2017 Workshop

We summarize the accomplishments of a multi-disciplinary workshop exploring the computational and scientific issues surrounding the discovery of linguistic units (subwords and words) in a language without orthography. We study the replacement of orthographic transcriptions by images and/or translated text in a well-resourced language to help unsupervised discovery from raw speech.

cs.CL↗

arXiv2015

Low-frequency conductivity in many-body localized systems

We argue that the a.c. conductivity $\sigma(\omega)$ in the many-body localized phase is a power law of frequency $\omega$ at low frequency: specifically, $\sigma(\omega) \sim \omega^\alpha$ with the exponent $\alpha$ approaching 1 at the phase transition to the thermal phase, and asymptoting to 2 deep in the localized phase. We identify two separate mechanisms giving rise to this power law: deep in the localized phase, the conductivity is dominated by rare resonant pairs of configurations; close to the transition, the dominant contributions are rare regions that are locally critical or in the thermal phase. We present numerical evidence supporting these claims, and discuss how these power laws can also be seen through polarization-decay measurements in ultracold atomic systems.

cond-mat.dis-nn↗

arXiv2014

Anomalous diffusion and Griffiths effects near the many-body localization transition

We explore the high-temperature dynamics of the disordered, one-dimensional XXZ model near the many-body localization (MBL) transition, focusing on the delocalized (i.e., "metallic") phase. In the vicinity of the transition, we find that this phase has the following properties: (i) Local magnetization fluctuations relax subdiffusively; (ii) the a.c. conductivity vanishes near zero frequency as a power law; (iii) the distribution of resistances becomes increasingly broad at low frequencies, approaching a power law in the zero-frequency limit. We argue that these effects can be understood in a unified way if the metallic phase near the MBL transition is a Griffiths phase. We establish scaling relations between the associated exponents, using exact linear-response arguments as well as a phenomenological resistor-capacitor model.

cond-mat.dis-nn↗

arXiv2014

Avalanches and hysteresis in frustrated superconductors and XY-spin-glasses

We study avalanches along the hysteresis loop of long-range interacting spin-glasses with continuous XY-symmetry - which serves as a toy model of granular superconductors with long-range and frustrated Josephson couplings. We identify sudden jumps in the $T=0$ configurations of the XY-phases, as an external field is increased. They are initiated by the softest mode of the inverse susceptibility matrix becoming unstable, which induces an avalanche of phase updates (or spin alignments). We analyze the statistics of these events, and study the correlation between the non-linear avalanches and the soft mode that initiates them. We find that the avalanches follow the directions of a small fraction of the softest modes of the inverse susceptibility matrix, similarly as was found in avalanches in jammed systems. In contrast to the similar Ising spin-glass (Sherrington-Kirkpatrick) studied previously, we find that avalanches are not distributed with a scale-free power law, but rather have a typical size which scales with the system size. We also observe that the Hessians of the spin-glass minima are not part of standard random matrix ensembles as the lowest eigenvector has a fractal support.

cond-mat.dis-nn↗

arXiv2014

Experimental Quantum Computations on a Topologically Encoded Qubit

The construction of a quantum computer remains a fundamental scientific and technological challenge, in particular due to unavoidable noise. Quantum states and operations can be protected from errors using protocols for fault-tolerant quantum computing (FTQC). Here we present a step towards this by implementing a quantum error correcting code, encoding one qubit in entangled states distributed over 7 trapped-ion qubits. We demonstrate the capability of the code to detect one bit flip, phase flip or a combined error of both, regardless on which of the qubits they occur. Furthermore, we apply combinations of the entire set of logical single-qubit Clifford gates on the encoded qubit to explore its computational capabilities. The implemented 7-qubit code is the first realization of a complete Calderbank-Shor-Steane (CSS) code and constitutes a central building block for FTQC schemes based on concatenated elementary quantum codes. It also represents the smallest fully functional instance of the color code, opening a route towards topological FTQC.

quant-ph↗

arXiv2013

Many-body Localization with Dipoles

Systems of strongly interacting dipoles offer an attractive platform to study many-body localized phases, owing to their long coherence times and strong interactions. We explore conditions under which such localized phases persist in the presence of power-law interactions and supplement our analytic treatment with numerical evidence of localized states in one dimension. We propose and analyze several experimental systems that can be used to observe and probe such states, including ultracold polar molecules and solid-state magnetic spin impurities.

cond-mat.stat-mech↗

arXiv2013

Localization of disordered bosons and magnets in random fields

We study localization properties of disordered bosons and spins in random fields at zero temperature. We focus on two representatives of different symmetry classes, hard-core bosons (XY magnets) and Ising magnets in random transverse fields, and contrast their physical properties. We describe localization properties using a locator expansion on general lattices. For 1d Ising chains, we find non-analytic behavior of the localization length as a function of energy at ω= 0, $ξ^{-1}(ω) = ξ^{-1}(0) + A |ω|^α$, with $α$ vanishing at criticality. This contrasts with the much smoother behavior predicted for XY magnets. We use these results to approach the ordering transition on Bethe lattices of large connectivity K, which mimic the limit of high dimensionality. In both models, in the paramagnetic phase with uniform disorder, the localization length is found to have a local maximum at ω= 0. For the Ising model, we find activated scaling at the phase transition, in agreement with infinite randomness studies. In the Ising model long range order is found to arise due to a delocalization and condensation initiated at ω= 0, without a closing mobility gap. We find that Ising systems establish order on much sparser (fractal) subgraphs than XY models. Possible implications of these results for finite-dimensional systems are discussed.

cond-mat.dis-nn↗

arXiv2013

Magnetoresistance and localization in bosonic insulators

We study the strong localization of hard core bosons. Using a locator expansion we find that in the insulator, unlike for typical fermion problems, nearly all low-energy scattering paths come with positive amplitudes and hence interfere constructively. As a consequence, the localization length of bosonic excitations shrinks when the constructive interference is suppressed by a magnetic field, entailing an exponentially large positive magnetoresistance, opposite to and significantly stronger than the analogous effect in fermions. Within the forward scattering approximation, we find that the lowest energy excitations are the most delocalized. A similar analysis applied to random field Ising models suggests that the ordering transition is due to a delocalization initiated at zero energy rather than due to the closure of a mobility gap in the paramagnet.

cond-mat.supr-con↗

arXiv2012

Magnetoresistance of an Anderson insulator of bosons

We study the magnetoresistance of two-dimensional bosonic Anderson insulators. We describe the change in spatial decay of localized excitations in response to a magnetic field, which is given by an interference sum over alternative tunnelling trajectories. The excitations become more localized with increasing field (in sharp contrast to generic fermionic excitations which get weakly delocalized): the localization length \xi(B) is found to change as \xi^{-1}(B)-\xi^{-1}(0)\sim B^{4/5}. The quantum interference problem maps onto the classical statistical mechanics of directed polymers in random media (DPRM). We explain the observed scaling using a simplified droplet model which incorporates the non-trivial DPRM exponents. Our results have implications for a variety of experiments on magnetic-field-tuned superconductor-to-insulator transitions observed in disordered films, granular superconductors, and Josephson junction arrays, as well as for cold atoms in artificial gauge fields.

cond-mat.dis-nn↗

arXiv2012

Quantum charge glasses of itinerant fermions with cavity-mediated long-range interactions

We study models of itinerant spinless fermions with random long-range interactions. We motivate such models from descriptions of fermionic atoms in multi-mode optical cavities. The solution of an infinite-range model yields a metallic phase which has glassy charge dynamics, and a localized glass phase with suppressed density of states at low energies. We compare these phases to the conventional disordered Fermi liquid, and the insulating electron glass of semiconductors. Prospects for the realization of such glassy phases in cold atom systems are discussed.

cond-mat.quant-gas↗

arXiv2012

Two-component Coulomb Glass in Disordered Superconducting Films

Motivated by evidence of local electron-electron attraction in experiments on disordered insulating films, we propose a new two-component Coulomb glass model that combines strong disorder and long-range Coulomb repulsion with the additional possibility of local pockets of a short-range inter-electron attraction. This model hosts a variety of interesting phenomena, in particular a crucial modification of the Coulomb gap previously believed to be universal. Tuning the short-range interaction to be repulsive, we find non-monotonic humps in the density of states within the Coulomb gap. We further study variable-range hopping transport in such systems by extending the standard resistor network approach to include the motion of both single electrons and local pairs. In certain parameter regimes the competition between these two types of carriers results in a distinct peak in resistance as a function of the local attraction strength, which can be tuned by a magnetic field.

cond-mat.dis-nn↗

arXiv2012

Collective excitations and marginal stability of quantum Ising spin glasses

We solve the Sherrington-Kirkpatrick (SK) model in a transverse field Γdeep in its quantum glass phase at zero temperature. We show that the glass phase is critical everywhere, exhibiting collective excitations with a gapless Ohmic spectral function. Using an effective potential approach, we interpret the latter as being due to disordered spin waves which behave as weakly coupled, underdamped harmonic oscillators. In the limit of small transverse field the low frequency tail of the spectrum tends to a universal, Γ-independent limit.

cond-mat.dis-nn↗