Search

The development of efficient surrogates of partial differential equations (PDEs) is a critical step towards scalable modeling of complex, multiscale systems-of-systems. Convolutional neural networks (CNNs) have gained popularity as the basis for such surrogate models due to their success in capturing high-dimensional input-output mappings and the negligible cost of a forward pass. However, the high cost of generating training data -- typically via classical numerical solvers -- raises the question of whether these models are worth pursuing over more straightforward alternatives with well-established theoretical foundations, such as Monte Carlo methods. To reduce the cost of data generation, we propose training a CNN surrogate model on a mixture of numerical solutions to both the $d$-dimensional problem and its ($d-1$)-dimensional approximation, taking advantage of the efficiency savings guaranteed by the curse of dimensionality. We demonstrate our approach on a multiphase flow test problem, using transfer learning to train a dense fully-convolutional encoder-decoder CNN on the two classes of data. Numerical results from a sample uncertainty quantification task demonstrate that our surrogate model outperforms Monte Carlo with several times the data generation budget.

Model merging aims to efficiently combine the weights of multiple expert models, each trained on a specific task, into a single multi-task model, with strong performance across all tasks. When applied to all but the last layer of weights, existing methods -- such as Task Arithmetic, TIES-merging, and TALL mask merging -- work well to combine expert models obtained by fine-tuning a common foundation model, operating within a "local" neighborhood of the foundation model. This work explores the more challenging scenario of "non-local" merging, which we find arises when an expert model changes significantly during pretraining or where the expert models do not even share a common foundation model. We observe that standard merging techniques often fail to generalize effectively in this non-local setting, even when accounting for permutation symmetries using standard techniques. We identify that this failure is, in part, due to "variance collapse", a phenomenon identified also in the setting of linear mode connectivity by Jordan et al. (2023). To address this, we propose a multi-task technique to re-scale and shift the output activations of the merged model for each task, aligning its output statistics with those of the corresponding task-specific expert models. Our experiments demonstrate that this correction significantly improves the performance of various model merging approaches in non-local settings, providing a strong baseline for future research on this problem.

DNA-Encoded Libraries (DEL) are combinatorial small molecule libraries that offer an efficient way to characterize diverse chemical spaces. Selection experiments using DELs are pivotal to drug discovery efforts, enabling high-throughput screens for hit finding. However, limited availability of public DEL datasets hinders the advancement of computational techniques designed to process such data. To bridge this gap, we present KinDEL, one of the first large, publicly available DEL datasets on two kinases: Mitogen-Activated Protein Kinase 14 (MAPK14) and Discoidin Domain Receptor Tyrosine Kinase 1 (DDR1). Interest in this data modality is growing due to its ability to generate extensive supervised chemical data that densely samples around select molecular structures. Demonstrating one such application of the data, we benchmark different machine learning techniques to develop predictive models for hit identification; in particular, we highlight recent structure-based probabilistic approaches. Finally, we provide biophysical assay data, both on- and off-DNA, to validate our models on a smaller subset of molecules. Data and code for our benchmarks can be found at: https://github.com/insitro/kindel.

Hydrological models often involve constitutive laws that may not be optimal in every application. We propose to replace such laws with the Kolmogorov-Arnold networks (KANs), a class of neural networks designed to identify symbolic expressions. We demonstrate KAN's potential on the problem of baseflow identification, a notoriously challenging task plagued by significant uncertainty. KAN-derived functional dependencies of the baseflow components on the aridity index outperform their original counterparts. On a test set, they increase the Nash-Sutcliffe Efficiency (NSE) by 67%, decrease the root mean squared error by 30%, and increase the Kling-Gupta efficiency by 24%. This superior performance is achieved while reducing the number of fitting parameters from three to two. Next, we use data from 378 catchments across the continental United States to refine the water-balance equation at the mean-annual scale. The KAN-derived equations based on the refined water balance outperform both the current aridity index model, with up to a 105% increase in NSE, and the KAN-derived equations based on the original water balance. While the performance of our model and tree-based machine learning methods is similar, KANs offer the advantage of simplicity and transparency and require no specific software or computational tools. This case study focuses on the aridity index formulation, but the approach is flexible and transferable to other hydrological processes.

Real-world systems are shaped by both their complex internal interactions and the changes in their noisy environments. In this work, we study how a shared active bath affects the statistical dependencies between two interacting Brownian particles by evaluating their mutual information. We decompose the mutual information into three terms: information stemming from the internal interactions between the particles; information induced by the shared bath, which encodes environmental changes; a term describing information interference that quantifies how the combined presence of both internal interactions and environment either masks (destructive interference) or boosts (constructive interference) information. By studying exactly the case of linear interactions, we find that the sign of information interference depends solely on that of the internal coupling. However, when internal interactions are described by a nonlinear activation function, we show that both constructive and destructive interference appear depending on the interplay between the timescale of the active environment, the internal interactions, and the environmental coupling. Finally, we show that our results generalize to hierarchical systems where asymmetric couplings to the environment mimic the scenario where the active bath is only partially accessible to one particle. This setting allows us to quantify how this asymmetry drives information interference. Our work underscores how information and functional relationships in complex multi-scale systems are fundamentally shaped by the environmental context.

We study the fundamental problem of sequential probability assignment, also known as online learning with logarithmic loss, with respect to an arbitrary, possibly nonparametric hypothesis class. Our goal is to obtain a complexity measure for the hypothesis class that characterizes the minimax regret and to determine a general, minimax optimal algorithm. Notably, the sequential $\ell_{\infty}$ entropy, extensively studied in the literature (Rakhlin and Sridharan, 2015, Bilodeau et al., 2020, Wu et al., 2023), was shown to not characterize minimax risk in general. Inspired by the seminal work of Shtarkov (1987) and Rakhlin, Sridharan, and Tewari (2010), we introduce a novel complexity measure, the \emph{contextual Shtarkov sum}, corresponding to the Shtarkov sum after projection onto a multiary context tree, and show that the worst case log contextual Shtarkov sum equals the minimax regret. Using the contextual Shtarkov sum, we derive the minimax optimal strategy, dubbed \emph{contextual Normalized Maximum Likelihood} (cNML). Our results hold for sequential experts, beyond binary labels, which are settings rarely considered in prior work. To illustrate the utility of this characterization, we provide a short proof of a new regret upper bound in terms of sequential $\ell_{\infty}$ entropy, unifying and sharpening state-of-the-art bounds by Bilodeau et al. (2020) and Wu et al. (2023)., Comment: To appear in NeurIPS 2024

While "complexity science" has achieved significant successes in several interdisciplinary fields such as economics and biology, it is only a very recent observation that legal systems -- from the way legal texts are drafted and connected to the rest of the corpus, up to the level of how judges and courts reach decisions under a variety of conflicting inputs -- share several features with standard Complex Adaptive Systems. This review is meant as a gentle introduction to the use of quantitative tools and techniques of complexity science to describe, analyse, and tame the complex web of human interactions that the Law is supposed to regulate. We offer an overview of the main directions of research undertaken so far as well as an outlook for future research, and we argue that statistical physicists and complexity scientists should not ignore the opportunities offered by the cross-fertilisation between legal scholarship and complex-systems modelling., Comment: 7 pages, 4 figures, 2 tables - submitted as EPL Perspective

Understanding the structural dynamics of many-particle glassy systems remains a key challenge in statistical physics. Over the last decade, glassy dynamics has also been reported in biological tissues, but is far from being understood. It was recently shown that vertex models of dense biological tissue exhibit very atypical, sub-Arrhenius dynamics, and here we ask whether such atypical structural dynamics of vertex models are related to unusual elastic properties. It is known that at zero temperature these models have an elasticity controlled by their under-constrained or isostatic nature, but little is known about how their elasticity varies with temperature. To address this question we investigate the 2D Voronoi model and measure the temperature dependence of the intermediate-time plateau shear modulus and the bulk modulus. We find that unlike in conventional glassformers, these moduli increase monotonically with temperature until the system fluidizes. We further show that the structural relaxation time can be quantitatively linked to the plateau shear modulus $G_p$, i.e.\ $G_p$ modulates the typical energy barrier scale for cell rearrangements. This suggests that the anomalous, structural dynamics of the 2D Voronoi model originates in its unusual elastic properties. Based on our results, we hypothesize that under-constrained systems might more generally give rise to a new class of "ultra-strong" glassformers., Comment: 5 pages, 5 figures

This paper presents a method for virtual contrast enhancement in breast MRI, offering a promising non-invasive alternative to traditional contrast agent-based DCE-MRI acquisition. Using a conditional generative adversarial network, we predict DCE-MRI images, including jointly-generated sequences of multiple corresponding DCE-MRI timepoints, from non-contrast-enhanced MRIs, enabling tumor localization and characterization without the associated health risks. Furthermore, we qualitatively and quantitatively evaluate the synthetic DCE-MRI images, proposing a multi-metric Scaled Aggregate Measure (SAMe), assessing their utility in a tumor segmentation downstream task, and conclude with an analysis of the temporal patterns in multi-sequence DCE-MRI generation. Our approach demonstrates promising results in generating realistic and useful DCE-MRI sequences, highlighting the potential of virtual contrast enhancement for improving breast cancer diagnosis and treatment, particularly for patients where contrast agent administration is contraindicated.

Text generation has a fundamental limitation almost by definition: there is no taking back tokens that have been generated, even when they are clearly problematic. In the context of language model safety, when a partial unsafe generation is produced, language models by their nature tend to happily keep on generating similarly unsafe additional text. This is in fact how safety alignment of frontier models gets circumvented in the wild, despite great efforts in improving their safety. Deviating from the paradigm of approaching safety alignment as prevention (decreasing the probability of harmful responses), we propose backtracking, a technique that allows language models to "undo" and recover from their own unsafe generation through the introduction of a special [RESET] token. Our method can be incorporated into either SFT or DPO training to optimize helpfulness and harmlessness. We show that models trained to backtrack are consistently safer than baseline models: backtracking Llama-3-8B is four times more safe than the baseline model (6.1\% $\to$ 1.5\%) in our evaluations without regression in helpfulness. Our method additionally provides protection against four adversarial attacks including an adaptive attack, despite not being trained to do so.

With close to three years of observations in hand, the NEID Earth Twin Survey (NETS) is starting to unearth new astrophysical signals for a curated sample of bright, radial velocity (RV)-quiet stars. We present the discovery of the first NETS exoplanet, HD 86728 b, a $m_p\sin i = 9.16^{+0.55}_{-0.56}\ \rm{M}_\oplus$ planet on a circular, $P=31.1503^{+0.0062}_{-0.0066}$ d orbit, thereby confirming a candidate signal identified by Hirsch et al. (2021). We confirm the planetary origin of the detected signal, which has a semi-amplitude of just $K=1.91^{+0.11}_{-0.12}$ m s$^{-1}$, via careful analysis of the NEID RVs and spectral activity indicators, and we constrain the mass and orbit via fits to NEID and archival RV measurements. The host star is intrinsically quiet at the $\sim1$ m s$^{-1}$ level, with the majority of this variability likely stemming from short-timescale granulation. HD 86728 b is among the small fraction of exoplanets with similar masses and periods that have no known planetary siblings., Comment: Submitted to AAS Journals. 18 pages, 10 figures, 3 tables, 1 appendix

Pre-trained deep learning embeddings have consistently shown superior performance over handcrafted acoustic features in speech emotion recognition (SER). However, unlike acoustic features with clear physical meaning, these embeddings lack clear interpretability. Explaining these embeddings is crucial for building trust in healthcare and security applications and advancing the scientific understanding of the acoustic information that is encoded in them. This paper proposes a modified probing approach to explain deep learning embeddings in the SER space. We predict interpretable acoustic features (e.g., f0, loudness) from (i) the complete set of embeddings and (ii) a subset of the embedding dimensions identified as most important for predicting each emotion. If the subset of the most important dimensions better predicts a given emotion than all dimensions and also predicts specific acoustic features more accurately, we infer those acoustic features are important for the embedding model for the given task. We conducted experiments using the WavLM embeddings and eGeMAPS acoustic features as audio representations, applying our method to the RAVDESS and SAVEE emotional speech datasets. Based on this evaluation, we demonstrate that Energy, Frequency, Spectral, and Temporal categories of acoustic features provide diminishing information to SER in that order, demonstrating the utility of the probing classifier method to relate embeddings to interpretable acoustic features.

We study the Berry curvature and Chern number of a non-collinear spin state on a honeycomb lattice that evolves from coplanar to ferromagnetic with a magnetic field applied along the $z$ axis. The coplanar state is stabilized by nearest-neighbor ferromagnetic interactions, single-ion anisotropy along $z$, and Dzyalloshinskii-Moriya interactions between next-nearest neighbor sites. Below the critical field $H_c$ that aligns the spins, the magnetic unit cell contains $M=6$ sites and the spin dynamics contains six magnon subbands. Although the classical energy is degenerate wrt the twist angle $\phi $ between nearest-neighbor spins, the dependence of the free energy on $\phi $ at low temperatures is dominated by the magnon zero-point energy, which contains extremum at $\phi =\pi l/3$ for integer $l$. The only unique ground states GS($\phi )$ have $l=0$ or 1. For $H < H_c'$, the zero-point energy has minima at even $l$ and the ground state is GS(0). For $H_c' < H < H_c$, the zero-point energy has minima at odd $l$ and the ground state is GS($\pi/3$). In GS(0), the magnon density-of-states exhibits five distinct phases with increasing field associated with the opening and closing of energy gaps between the two or three magnonic bands, each containing between 1 and 4 four magnon subbands. While the Berry curvature vanishes for the coplanar $\phi=0$ phase in zero field, the Berry curvature and Chern numbers exhibit signatures of the five phases at nonzero fields below $H_c'$. If $\phi \ne \pi l/3$, the Chern numbers of the two or three magnonic bands are non-integer. We also evaluate the inelastic neutron-scattering spectrum $S(\vk ,\omega )$ produced by the six magnon subbands in all five phases of GS(0) and in GS($\pi/3$)., Comment: 13 pages, 12 figures

When a millimetric body is placed atop a vibrating liquid bath, the relative motion between the object and interface generates outward propagating waves with an associated momentum flux. Prior work has shown that isolated chiral objects, referred to as spinners, can thus rotate steadily in response to their self-generated wavefield. Here, we consider the case of two co-chiral spinners held at a fixed spacing from one another but otherwise free to interact hydrodynamically through their shared fluid substrate. Two identical spinners are able to synchronize their rotation, with their equilibrium phase difference sensitive to their spacing and initial conditions, and even cease to rotate when the coupling becomes sufficiently strong. Non-identical spinners can also find synchrony provided their intrinsic differences are not too disparate. A hydrodynamic wave model of the spinner interaction is proposed, recovering all salient features of the experiment. In all cases, the spatially periodic nature of the capillary wave coupling is directly reflected in the emergent equilibrium behaviors.

We develop variational search distributions (VSD), a method for finding discrete, combinatorial designs of a rare desired class in a batch sequential manner with a fixed experimental budget. We formalize the requirements and desiderata for this problem and formulate a solution via variational inference. In particular, VSD uses off-the-shelf gradient based optimization routines, can learn powerful generative models for designs, and can take advantage of scalable predictive models. We derive asymptotic convergence rates for learning the true conditional generative distribution of designs with certain configurations of our method. After illustrating the generative model on images, we empirically demonstrate that VSD can outperform existing baseline methods on a set of real sequence-design problems in various biological systems., Comment: 34 pages with supplementary material included

Flocking phase transitions found in models of polar active matter are paradigmatic examples of active phase transitions in soft matter. An interesting specialization of flocking models concerns a ``topological'' vs ``metric'' choice by which agents are considered to be interacting neighbors. While recent theoretical work suggests that the order-disorder transition in these polar aligning models is universally first order, numerical studies have suggested that topological models may instead have a continuous transition. Some recent simulations have found that some variations of topologically interacting flocking agents have a discontinuous transition, but unambiguous observations of phase coexistence using common Voronoi-based alignment remains elusive. In this work, we use a custom GPU-accelerated simulation package to perform million-particle-scale simulations of these Voronoi-Vicsek flocking models. By accessing such large systems on appropriately long time scales, we are able to show that a regime of stable phase coexistence between the ordered and disordered phases, confirming the discontinuous nature of this transition in the thermodynamic limit., Comment: 7 pages, 7 figures, many simulations

The vertical transport of solid material in a stratified medium is fundamental to a number of environmental applications, with implications for the carbon cycle and nutrient transport in marine ecosystems. In this work, we study the diffusion-limited settling of highly porous particles in a density-stratified fluid through a combination of experiment, analysis, and numerical simulation. By delineating and appealing to the diffusion-limited regime wherein buoyancy effects due to mass adaptation dominate hydrodynamic drag, we derive a simple expression for the steady settling velocity of a sphere as a function of the density, size, and diffusivity of the solid, as well as the density gradient of the background fluid. In this regime, smaller particles settle faster, in contrast with most conventional hydrodynamic drag mechanisms. Furthermore, we outline a general mathematical framework for computing the steady settling speed of a body of arbitrary shape in this regime and compute exact results for the case of general ellipsoids. Using hydrogels as a highly porous model system, we validate the predictions with laboratory experiments in linear stratification for a wide range of parameters. Lastly, we show how the predictions can be applied to arbitrary slowly varying background density profiles and demonstrate how a measured particle position over time can be used to reconstruct the background density profile.

Observations of glassy dynamics in experiments on confluent cellular tissue have inspired a wealth of computational and theoretical research to model their emergent collective behavior. Initial studies of the physical properties of several geometric cell models, including vertex-type models, have highlighted anomalous sub-Arrhenius, or "ultra-strong," scaling of the dynamics with temperature. Here we show that the dynamics and material properties of the 2d Voronoi model deviate even further from the standard glassforming paradigm. By varying the characteristic shape index $p_0$, we demonstrate that the system properties can be tuned between displaying expected glassforming behavior, including the breakdown of the Stokes-Einstein-Sutherland relation and the formation of dynamical heterogeneities, and an unusual regime in which the viscosity does not diverge as the characteristic relaxation time increase and dynamical heterogeneities are strongly suppressed. Our results provide further insight into the fundamental properties of this class of anomalous glassy materials, and provide a step towards designing materials with predetermined glassy dynamics., Comment: 10 pages, 4 figures

We study the task of testable learning of general -- not necessarily homogeneous -- halfspaces with adversarial label noise with respect to the Gaussian distribution. In the testable learning framework, the goal is to develop a tester-learner such that if the data passes the tester, then one can trust the output of the robust learner on the data.Our main result is the first polynomial time tester-learner for general halfspaces that achieves dimension-independent misclassification error. At the heart of our approach is a new methodology to reduce testable learning of general halfspaces to testable learning of nearly homogeneous halfspaces that may be of broader interest., Comment: Presented to COLT'24

A $(v,k,\lambda)$-difference set in a group $G$ of order $v$ is a subset $\{d_1, d_2, \ldots,d_k\}$ of $G$ such that $D=\sum d_i$ in the group ring ${\mathbb Z}[G]$ satisfies $$D D^{-1} = n + \lambda G,$$ where $n=k-\lambda$. In other words, the nonzero elements of $G$ all occur exactly $\lambda$ times as differences of elements in $D$. A $(v,k,\lambda,t)$-almost difference set has $t$ nonzero elements of $G$ occurring $\lambda$ times, and the other $v-1-t$ occurring $\lambda+1$ times. When $\lambda=0$, this is equivalent to a modular Golomb ruler. In this paper we investigate existence questions on these objects, and extend previous results constructing almost difference sets by adding or removing an element from a difference set., Comment: 6 pages, 1 figure

With the growing demand for higher wireless data rates, the interest in extending the carrier frequency of wireless links to the terahertz (THz) range has significantly increased. For long-distance outdoor wireless communications, THz channels may suffer substantial power loss and security issues due to atmospheric weather effects. It is crucial to assess the impact of weather on high-capacity data transmission to evaluate wireless system link budgets and performance accurately. In this article, we provide an insight into the propagation characteristics of THz channels under atmospheric conditions and the security aspects of THz communication systems in future applications. We conduct a comprehensive survey of our recent research and experimental findings on THz channel transmission and physical layer security, synthesizing and categorizing the state-of-the-art research in this domain. Our analysis encompasses various atmospheric phenomena, including molecular absorption, scattering effects, and turbulence, elucidating their intricate interactions with THz waves and the resultant implications for channel modeling and system design. Furthermore, we investigate the unique security challenges posed by THz communications, examining potential vulnerabilities and proposing novel countermeasures to enhance the resilience of these high-frequency systems against eavesdropping and other security threats. Finally, we discuss the challenges and limitations of such high-frequency wireless communications and provide insights into future research prospects for realizing the 6G vision, emphasizing the need for innovative solutions to overcome the atmospheric hurdles and security concerns in THz communications., Comment: Submitted to Fundamental Research

Quantum resonances in collisions and reactions are a sensitive probe of the intermolecular forces. They may dominate the final quantum state distribution, as recently observed for Feshbach resonances in a cold collision experiment (Science 380, 77 (2023)). This raises the question whether the sensitivity of such measurements is sufficient to assess the quality of theoretical models for the interaction. We here compare measured collision cross sections to those obtained with exact quantum coupled-channels scattering calculations for three different ab initio potential energy surfaces. We find that the ability to test the correct prediction of energy redistribution over molecular degrees of freedom is within reach, requiring only a modest improvement in energy resolution of current experiments. Such improvement will enable the separation of individual resonances and allow for an unambiguous experimental test of different theory approaches.

The helium atom, with one nucleus and two electrons, is a prototypical system to study quantum many-body dynamics. Doubly-excited states, or quantum states in which both electrons are excited by one photon, are showcase scenarios of electronic-correlation mediated effects. In this paper, the natural lifetimes of the doubly-excited $^1$P$^o$ 2s$n$p Rydberg series and the $^1$S$^e$ 2p$^2$ dark state in helium in the 60 eV to 65 eV region are measured directly in the time domain with extreme-ultraviolet/near-infrared noncollinear attosecond four-wave-mixing (FWM) spectroscopy. The measured lifetimes are in agreement with lifetimes deduced from spectral linewidths and theoretical predictions, and the roles of specific decay mechanisms are considered. While complex spectral line shapes in the form of Fano resonances are common in absorption spectroscopy of autoionizing states, the background-free and thus homodyned character of noncollinear FWM results exclusively in Lorentzian spectral features in the absence of strong-field effects. The onset of strong-field effects that would affect the extraction of accurate natural lifetimes in helium by FWM is determined to be approximately 0.3 Rabi cycles. This study provides a systematic understanding of the FWM parameters necessary to enable accurate lifetime extractions, which can be utilized in more complex quantum systems such as molecules in the future., Comment: 10 pages, 5 figures, 1 table. P. R. and N. G. P. contributed equally to this work

The propagation of millimeter wave (MMW) and sub-terahertz (THz) signals through plasma sheaths is a critical concern for maintaining communication with hypersonic vehicles, yet the impact of complex plasma structures on these high-frequency channels remains insufficiently understood. In this work, we aim to characterize the transmission properties of MMW and sub-THz waves through plasma sheaths with various density profiles and ripple structures, addressing the gap in knowledge regarding the effects of plasma inhomogeneities on signal propagation. We employ an approach combining Inductively Coupled Plasma (ICP) data with transfer matrix methods (TMM) to model propagation through both flat and rippled plasma layers. Our findings reveal that ripple structures in plasma sheaths significantly affect channel performance, with periodic ripples reducing cutoff frequency and introducing frequency-selective behavior, while random ripples cause more unpredictable transmission characteristics. Our results explore the impact of the arrangement of plasma density layers and the parameters of ripple structures (period and amplitude) on channel transmission, group velocity dispersion, and angular dependence of wave propagation. These results provide crucial insights for the design and optimization of communication systems for hypersonic vehicles, potentially enabling the development of adaptive technologies capable of maintaining reliable communication in complex plasma environments., Comment: Submitted to IEEE Transactions on Plasma Science

Topological insulators offer unique opportunities for novel electronics and quantum phenomena. However, intrinsic material limitations often restrict their applications and practical implementation. Over a decade ago it was predicted that a time-periodic perturbation can generate out-of-equilibrium states known as Floquet topological insulators (FTIs), hosting topologically protected transport and anomalous Hall physics, and opening routes to optically tunable bandstructures and devices compatible with petahertz electronics. Although such states have not yet been directly observed, indirect signatures such as the light-induced anomalous Hall effect were recently measured. Thus far, much remained experimentally unclear and fundamentally unknown about solid-state FTI and whether they can be employed for electronics. Here we demonstrate coherent control of photocurrents in light-dressed graphene. Circularly-polarized laser pulses dress the graphene band structure to obtain an FTI, and phase-locked second harmonic pulses drive electrons in the FTI. This approach allows us to measure resulting all-optical anomalous Hall photocurrents, FTI-valley-polarized currents, and photocurrent circular dichroism, all phenomena that put FTIs on equal footing with equilibrium topological insulators. We further present an intuitive description for the sub-optical-cycle light-matter interaction, revealing dynamical symmetry selection rules for photocurrents. All measurements are supported by strong agreement with ab-initio and analytic theory. Remarkably, the photocurrents show a strong sub-cycle phase-sensitivity that can be employed for ultrafast control in topotronics and spectroscopy. Our work connects Floquet and topological physics with attoscience and valleytronics, and goes beyond band structure engineering by initiating lightwave-driven dynamics in FTI states., Comment: 14 pages, 8 figures

In molecular physics, it is often necessary to average over the orientation of molecules when calculating observables, in particular when modelling experiments in the liquid or gas phase. Evaluated in terms of Euler angles, this is closely related to integration over two- or three-dimensional unit spheres, a common problem discussed in numerical analysis. The computational cost of the integration depends significantly on the quadrature method, making the selection of an appropriate method crucial for the feasibility of simulations. After reviewing several classes of spherical quadrature methods in terms of their efficiency and error distribution, we derive guidelines for choosing the best quadrature method for orientation averages and illustrate these with three examples from chiral molecule physics. While Gauss quadratures allow for achieving numerically exact integration for a wide range of applications, other methods offer advantages in specific circumstances. Our guidelines can also by applied to higher-dimensional spherical domains and other geometries. We also present a Python package providing a flexible interface to a variety of quadrature methods.

Recent studies have underscored the capabilities of natural imaging foundation models to serve as powerful feature extractors, even in a zero-shot setting for medical imaging data. Most commonly, a shallow multi-layer perceptron (MLP) is appended to the feature extractor to facilitate end-to-end learning and downstream prediction tasks such as classification, thus representing the de facto standard. However, as graph neural networks (GNNs) have become a practicable choice for various tasks in medical research in the recent past, we direct attention to the question of how effective GNNs are compared to MLP prediction heads for the task of 3D medical image classification, proposing them as a potential alternative. In our experiments, we devise a subject-level graph for each volumetric dataset instance. Therein latent representations of all slices in the volume, encoded through a DINOv2 pretrained vision transformer (ViT), constitute the nodes and their respective node features. We use public datasets to compare the classification heads numerically and evaluate various graph construction and graph convolution methods in our experiments. Our findings show enhancements of the GNN in classification performance and substantial improvements in runtime compared to an MLP prediction head. Additional robustness evaluations further validate the promising performance of the GNN, promoting them as a suitable alternative to traditional MLP classification heads. Our code is publicly available at: https://github.com/compai-lab/2024-miccai-grail-kiechle, Comment: Accepted at MICCAI 2024 - GRAIL Workshop

The discovery of small organic compounds for inducing stem cell differentiation is a time- and resource-intensive process. While data science could, in principle, facilitate the discovery of these compounds, novel approaches are required due to the difficulty of acquiring training data from large numbers of example compounds. In this paper, we demonstrate the design of a new compound for inducing cardiomyocyte differentiation using simple regression models trained with a data set containing only 80 examples. We introduce decorated shape descriptors, an information-rich molecular feature representation that integrates both molecular shape and hydrophilicity information. These models demonstrate improved performance compared to ones using standard molecular descriptors based on shape alone. Model overtraining is diagnosed using a new type of sensitivity analysis. Our new compound is designed using a conservative molecular design strategy, and its effectiveness is confirmed through expression profiles of cardiomyocyte-related marker genes using real-time polymerase chain reaction experiments on human iPS cell lines. This work demonstrates a viable data-driven strategy for designing new compounds for stem cell differentiation protocols and will be useful in situations where training data is limited., Comment: 26 pages (main paper), including 7 figures and 3 tables. 23 pages of supporting information. To be submitted to a journal

Denoising diffusion probabilistic models enable high-fidelity image synthesis and editing. In biomedicine, these models facilitate counterfactual image editing, producing pairs of images where one is edited to simulate hypothetical conditions. For example, they can model the progression of specific diseases, such as stroke lesions. However, current image editing techniques often fail to generate realistic biomedical counterfactuals, either by inadequately modeling indirect pathological effects like brain atrophy or by excessively altering the scan, which disrupts correspondence to the original images. Here, we propose MedEdit, a conditional diffusion model for medical image editing. MedEdit induces pathology in specific areas while balancing the modeling of disease effects and preserving the integrity of the original scan. We evaluated MedEdit on the Atlas v2.0 stroke dataset using Frechet Inception Distance and Dice scores, outperforming state-of-the-art diffusion-based methods such as Palette (by 45%) and SDEdit (by 61%). Additionally, clinical evaluations by a board-certified neuroradiologist confirmed that MedEdit generated realistic stroke scans indistinguishable from real ones. We believe this work will enable counterfactual image editing research to further advance the development of realistic and clinically useful imaging tools., Comment: Accepted at MICCAI24 Simulation and Synthesis in Medical Imaging (SASHIMI) workshop

Deep learning holds immense promise for aiding radiologists in breast cancer detection. However, achieving optimal model performance is hampered by limitations in availability and sharing of data commonly associated to patient privacy concerns. Such concerns are further exacerbated, as traditional deep learning models can inadvertently leak sensitive training information. This work addresses these challenges exploring and quantifying the utility of privacy-preserving deep learning techniques, concretely, (i) differentially private stochastic gradient descent (DP-SGD) and (ii) fully synthetic training data generated by our proposed malignancy-conditioned generative adversarial network. We assess these methods via downstream malignancy classification of mammography masses using a transformer model. Our experimental results depict that synthetic data augmentation can improve privacy-utility tradeoffs in differentially private model training. Further, model pretraining on synthetic data achieves remarkable performance, which can be further increased with DP-SGD fine-tuning across all privacy guarantees. With this first in-depth exploration of privacy-preserving deep learning in breast imaging, we address current and emerging clinical privacy requirements and pave the way towards the adoption of private high-utility deep diagnostic models. Our reproducible codebase is publicly available at https://github.com/RichardObi/mammo_dp., Comment: Early Accept at MICCAI 2024 Deep-Breath Workshop

In this work, we introduce Progressive Growing of Patch Size, a resource-efficient implicit curriculum learning approach for dense prediction tasks. Our curriculum approach is defined by growing the patch size during model training, which gradually increases the task's difficulty. We integrated our curriculum into the nnU-Net framework and evaluated the methodology on all 10 tasks of the Medical Segmentation Decathlon. With our approach, we are able to substantially reduce runtime, computational costs, and CO2 emissions of network training compared to classical constant patch size training. In our experiments, the curriculum approach resulted in improved convergence. We are able to outperform standard nnU-Net training, which is trained with constant patch size, in terms of Dice Score on 7 out of 10 MSD tasks while only spending roughly 50% of the original training runtime. To the best of our knowledge, our Progressive Growing of Patch Size is the first successful employment of a sample-length curriculum in the form of patch size in the field of computer vision. Our code is publicly available at https://github.com/compai-lab/2024-miccai-fischer., Comment: Accepted at MICCAI2024; Changes for Camera-Ready-Version for MICCAI2024 (missing in this arxiv submission): Replaced T-Test with Wilcoxon Signed Ranked Test, as DSC samples are not normally distributed => now only significant improvements and no significant decreases in performance for PGPS/PGPS+

F. Luca proved for any fixed rational number $\alpha>0$ that the Diophantine equations of the form $\alpha\,m!=f(n!)$, where $f$ is either the Euler function or the divisor sum function or the function counting the number of divisors, have only finitely many integer solutions $(m,n)$. In this paper we generalize the mentioned result and show that Diophantine equations of the form $\alpha\,m_1!\cdots m_r!=f(n!)$ have finitely many integer solutions, too. In addition, we do so by including the case $f$ is the sum of $k$\textsuperscript{th} powers of divisors function. Moreover, we observe that the same holds by replacing some of the factorials with certain examples of Bhargava factorials., Comment: 8 pages, comments welcome!

The properties of tissue interfaces -- between separate populations of cells, or between a group of cells and its environment -- has attracted intense theoretical, computational, and experimental study. Recent work on shape-based models inspired by dense epithelia have suggested a possible ``topological sharpening'' effect, by which four-fold vertices spatially coordinated along a cellular interface lead to a cusp-like restoring force acting on cells at the interface, which in turn greatly suppresses interfacial fluctuations. We revisit these interfacial fluctuations, focusing on the distinction between short length scale reduction of interfacial fluctuations and long length scale renormalized surface tension. To do this, we implement a spectrally resolved analysis of fluctuations over extremely long simulation times. This leads to more quantitative information on the topological sharpening effect, in which the degree of sharpening depends on the length scale over which it is measured. We compare our findings with a Brownian bridge model of the interface, and close by analyzing existing experimental data in support of the role of short-length-scale topological sharpening effects in real biological systems., Comment: 9 pages, 7 figures

In this work, we investigate the problem of adapting to the presence or absence of causal structure in multi-armed bandit problems. In addition to the usual reward signal, we assume the learner has access to additional variables, observed in each round after acting. When these variables $d$-separate the action from the reward, existing work in causal bandits demonstrates that one can achieve strictly better (minimax) rates of regret (Lu et al., 2020). Our goal is to adapt to this favorable "conditionally benign" structure, if it is present in the environment, while simultaneously recovering worst-case minimax regret, if it is not. Notably, the learner has no prior knowledge of whether the favorable structure holds. In this paper, we establish the Pareto optimal frontier of adaptive rates. We prove upper and matching lower bounds on the possible trade-offs in the performance of learning in conditionally benign and arbitrary environments, resolving an open question raised by Bilodeau et al. (2022). Furthermore, we are the first to obtain instance-dependent bounds for causal bandits, by reducing the problem to the linear bandit setting. Finally, we examine the common assumption that the marginal distributions of the post-action contexts are known and show that a nontrivial estimate is necessary for better-than-worst-case minimax rates., Comment: Accepted to ICML 2024

One of the hypothesized functions of biomolecular condensates is to act as chemical reactors, where chemical reactions can be modulated, i.e. accelerated or slowed down, while substrate molecules enter and products exit from the condensate. Likewise, the components themselves that take part in the architectural integrity of condensates might be modified by active (energy consuming, non-equilibrium) processes, e.g. by ATPase chaperones or by kinases and phosphatases. In this work, we study how the presence of spatial inhomogeneities, such as in the case of liquid-liquid phase separation, affects active chemical reactions and results in the presence of directional flows of matter, which are one of the hallmarks of non-equilibirum processes. We establish the minimal conditions for the existence of such spatial currents, and we furthermore find that these fluxes are maximal at the condensate interface. These results propose that some condensates might be most efficient as chemical factories due to their interfaces rather than their volumes, and could suggest a possible biological reason for the the observed abundance of small non-fusing condensates inside the cell, thus maximizing their surface and the associated fluxes., Comment: 15 pages, 4 figures

One of the most significant problems in cuneiform pedagogy is the process of looking up unknown signs, which often involves a tedious page-by-page search through a sign list. This paper proposes a new "recursive encoding" for signs, which represents the arrangement of strokes in a way a computer can process. A series of new algorithms then offers students a new way to look up signs by any distinctive component, as well as providing new ways to render signs and tablets electronically., Comment: 27 pages, 29 figures, 5 tables

Conventional wisdom dictates that quantum effects become unimportant at high temperatures. In magnets, when the thermal energy exceeds interactions between atomic magnetic moments, the moments are usually uncorrelated, and classical paramagnetic behavior is observed. This thermal decoherence of quantum spin behaviors is a major hindrance to quantum information applications of spin systems. Remarkably, our neutron scattering experiments on Yb chains in an insulating perovskite crystal defy these conventional expectations. We find a sharply defined spectrum of spinons, fractional quantum excitations of spin-1/2 chains, to persist to temperatures much higher than the scale of the interactions between Yb magnetic moments. The observed sharpness of the spinon continuum's dispersive upper boundary indicates a spinon mean free path exceeding $\approx 35$ inter-atomic spacings at temperatures more than an order of magnitude above the interaction energy scale. We thus discover an important and highly unique quantum behavior, which expands the realm of quantumness to high temperatures where entropy-governed classical behaviors were previously believed to dominate. Our results have profound implications for spin systems in quantum information applications operating at finite temperatures and motivate new developments in quantum metrology., Comment: 17 pages, 4 figures main text plus 19 pages, 7 figures supplementary text

Pathological lymph node delineation is crucial in cancer diagnosis, progression assessment, and treatment planning. The MICCAI 2023 Lymph Node Quantification Challenge published the first public dataset for pathological lymph node segmentation in the mediastinum. As lymph node annotations are expensive, the challenge was formed as a weakly supervised learning task, where only a subset of all lymph nodes in the training set have been annotated. For the challenge submission, multiple methods for training on these weakly supervised data were explored, including noisy label training, loss masking of unlabeled data, and an approach that integrated the TotalSegmentator toolbox as a form of pseudo labeling in order to reduce the number of unknown voxels. Furthermore, multiple public TCIA datasets were incorporated into the training to improve the performance of the deep learning model. Our submitted model achieved a Dice score of 0.628 and an average symmetric surface distance of 5.8~mm on the challenge test set. With our submitted model, we accomplished third rank in the MICCAI2023 LNQ challenge. A finding of our analysis was that the integration of all visible, including non-pathological, lymph nodes improved the overall segmentation performance on pathological lymph nodes of the test set. Furthermore, segmentation models trained only on clinically enlarged lymph nodes, as given in the challenge scenario, could not generalize to smaller pathological lymph nodes. The code and model for the challenge submission are available at \url{https://gitlab.lrz.de/compai/MediastinalLymphNodeSegmentation}., Comment: Accepted for publication at the Journal of Machine Learning for Biomedical Imaging (MELBA) https://melba-journal.org/2024:008

We study how to optimally segment a monopolistic market given a redistributive objective. Optimal redistributive segmentations (i) induce the seller to price progressively, i.e., richer consumers pay higher prices than poorer ones, and (ii) may require giving a higher profit than uniform pricing if the redistributive motive is strong. We further show that optimal redistributive segmentations are implementable via price-based regulation.

Human breath contains over 3000 volatile organic compounds, abnormal concentrations of which can indicate the presence of certain diseases. Recently, metal-organic framework (MOF)-metal oxide composite materials have been explored for chemiresistive sensor applications, however their ability to detect breath compounds associated with specific diseases remains unknown. In this work, we present a new high-throughput computational protocol for evaluating the sensing ability of MOF-metal oxide towards small organic compounds. This protocol uses a cluster-based method for accelerated structure relaxation, and a combination of binding energies and density-of-states analysis to evaluate sensing ability, the latter measured using Wasserstein distances. We apply this protocol to the case of the MOF-metal oxide composite material NM125-TiO2 and show that it is consistent with previously reported experimental results for this system. We examine the sensing ability of NM125-TiO2 for over 100 human-breath compounds spanning 13 different diseases. Statistical inference then allows us to identifies ones which subsequent experimental efforts should focus on. Overall, this work provides new tools for computational sensor research, while also illustrating how computational materials science can be integrated into the field of preventative medicine., Comment: 20 pages with 22 pages of supporting information. 7 figures and 1 table. In preparation for submission to a journal

Commercially impactful quantum algorithms such as quantum chemistry and Shor's algorithm require a number of qubits and gates far beyond the capacity of any existing quantum processor. Distributed architectures, which scale horizontally by networking modules, provide a route to commercial utility and will eventually surpass the capability of any single quantum computing module. Such processors consume remote entanglement distributed between modules to realize distributed quantum logic. Networked quantum computers will therefore require the capability to rapidly distribute high fidelity entanglement between modules. Here we present preliminary demonstrations of some key distributed quantum computing protocols on silicon T centres in isotopically-enriched silicon. We demonstrate the distribution of entanglement between modules and consume it to apply a teleported gate sequence, establishing a proof-of-concept for T centres as a distributed quantum computing and networking platform., Comment: 14 pages, 13 figures

A codimension-two spiral spin-liquid is a correlated paramagnetic state with one-dimensional ground state degeneracy hosted within a three-dimensional lattice. Here, via neutron scattering experiments and numerical simulations, we establish the existence of a codimension-two spiral spin-liquid in the effective honeycomb-lattice compound Cs$_3$Fe$_2$Cl$_9$ and demonstrate the selective visibility of the spiral surface through phase tuning. In the long-range ordered regime, competing spiral and spin density wave orders emerge as a function of applied magnetic field, among which a possible order-by-disorder transition is identified., Comment: 21 pages, 23 figures