Search

We derive a class of multi-species aggregation-diffusion systems from stochastic interacting particle systems via relative entropy method with quantitative bounds. We show an algebraic $L^1$-convergence result using moderately interacting particle systems approximating attractive/repulsive singular potentials up to Newtonian/Coulomb singularities without additional cut-off on the particle level. The first step is to make use of the relative entropy between the joint distribution of the particle system and an approximated limiting aggregation-diffusion system. A crucial argument in the proof is to show convergence in probability by a stopping time argument. The second step is to obtain a quantitative convergence rate to the limiting aggregation-diffusion system from the approximated PDE system. This is shown by evaluating a combination of relative entropy and $L^2$-distance., Comment: 43 pages

We revisit the Dicke superradiance problem, namely finding the time evolution of the density operator for an initially inverted ensemble of $N$ identical two-level systems (spins), undergoing collective spontaneous decay. We derive a compactly written, easily computable and fully analytical solution, valid for any time during the evolution and for any $N$.

We propose an efficient method to numerically simulate the dissipative dynamics of large numbers of quantum emitters in ordered arrays in the presence of long-range dipole-dipole interactions mediated by the vacuum electromagnetic field. Using the spatial symmetries of the system, we rewrite the equations of motion in a collective spin basis and subsequently apply a higher-order cumulant expansion for the collective operators. By truncating the subradiant collective modes with a heavily suppressed decay rate and keeping only the effect from the radiating collective modes, we reduce the numerical complexity significantly. This allows to efficiently compute the dissipative dynamics of the observables of interest for a linear, ring-shaped and planar arrays of quantum emitters. In particular, we characterize the excited population, the total photon emission rate and the second order intensity correlation function $g^{(2)}(\tau =0)$, which are challenging to compute for large systems with traditional cumulant expansion methods based on the individual spins., Comment: 19 pages, 8 figures

Subwavelength ring-shaped structures of quantum emitters exhibit outstanding radiation properties and are useful for antennas, excitation transport, and storage. Taking inspiration from the oligomeric geometry of biological light-harvesting 2 (LH2) complexes, we study here generic examples and predict highly efficient excitation transfer in a three-dimensional (3D) subwavelength concentric stacked ring structure with a diameter of 400 $nm$, formed by two-level atoms. Utilizing the quantum optical open system master equation approach for the collective dipole dynamics, we demonstrate that, depending on the system parameters, our bio-mimicked 3D ring enables efficient excitation transfer between two ring layers. Our findings open prospects for engineering other biomimetic light-matter platforms and emitter arrays to achieve efficient energy transfer., Comment: Minor revision, references added, 15 pages, 10 figures, 6 tables

Subwavelength atomic arrays in free space are becoming a leading platform for exploring emergent many-body quantum phenomena. These arrays feature strong light-induced dipole-dipole interactions, resulting in subradiant collective resonances characterized by narrowed linewidths. In this work, we present a sideband cooling scheme for atoms trapped in subwavelength arrays that utilizes these narrow collective resonances. We derive an effective master equation for the atomic motion by adiabatically eliminating the internal degrees of freedom of the atoms, and validate its prediction with numerical simulations of the full system. Our results demonstrate that subradiant resonances enable the cooling of ensembles of atoms to temperatures lower than those achievable without dipole interactions, provided the atoms have different trap frequencies. Remarkably, narrow collective resonances can be sideband-resolved even when the individual atomic transition is not. In such scenarios, ground state cooling becomes feasible solely due to light-induced dipole-dipole interactions. This approach could be utilized for future quantum technologies based on dense ensembles of emitters, and paves the way towards harnessing many-body cooperative decay for enhanced motional control., Comment: 10 pages of main text and appendix, with 4 figures. 9 pages of Supplemental Material, with 4 figures

A central limit theorem is shown for moderately interacting particles in the whole space. The interaction potential approximates singular attractive or repulsive potentials of sub-Coulomb type. It is proved that the fluctuations become asymptotically Gaussians in the limit of infinitely many particles. The methodology is inspired by the classical work of Oelschl\"ager on fluctuations for the porous-medium equation. The novelty in this work is that we can allow for attractive potentials in the moderate regime and still obtain asymptotic Gaussian fluctuations. The key element of the proof is the mean-square convergence in expectation for smoothed empirical measures associated to moderately interacting $N$-particle systems with rate $N^{-1/2-\varepsilon}$ for some $\varepsilon>0$. To allow for attractive potentials, the proof uses a quantitative mean-field convergence in probability with any algebraic rate and a law-of-large-numbers estimate as well as a systematic separation of the terms to be estimated in a mean-field part and a law-of-large-numbers part.

Non-adiabatic molecular phenomena, arising from the breakdown of the Born-Oppenheimer approximation, govern the fate of virtually all photo-physical and photochemical processes and limit the quantum efficiency of molecules and other solid-state embedded quantum emitters. A simple and elegant description, the energy gap law, was derived five decades ago, predicting that the non-adiabatic coupling between the excited and ground potential landscapes lead to non-radiative decay with a quasi-exponential dependence on the energy gap. We revisit and extend this theory to account for crucial aspects such as vibrational relaxation, dephasing, and radiative loss. We find a closed analytical solution with general validity which indicates a direct proportionality of the non-radiative rate with the vibrational relaxation rate at low temperatures, and with the dephasing rate of the electronic transition at high temperatures. Our work establishes a connection between nanoscale quantum optics, open quantum system dynamics and non-adiabatic molecular physics., Comment: 18 pages and 5 figures

The emergence of the COVID-19 pandemic, followed by policy measures to combat the virus, evoked public protest movements world-wide. These movements emerged through virtual social networks as well as local protest gatherings. Prior research has studied such movements solely in the virtual space through social network analysis, thereby disregarding the role of local interaction for protest. This study, however, recognizes the importance of the geo-spatial dimension in protest movements. We therefore introduce a large-scale spatial-social network analysis of a georeferenced Twitter user network to understand the regional connections and transnational influences of the Austrian COVID-19 protest movement through the social network. Our findings reveal that the virtual network is distinctly structured along geographic and linguistic boundaries. We further find that the movement is clearly organized along national protest communities. These results highlight the importance of regional and local influencing factors over the impact of transnational influences for the protest movement.

Light-harvesting via the transport and trapping of optically-induced electronic excitations is of fundamental interest to the design of new energy efficient quantum technologies. Using a paradigmatic quantum optical model, we study the influence of coherence, entanglement, and cooperative dissipation on the transport and capture of excitation energy. In particular, we demonstrate that the rate of energy extraction is optimized under conditions that minimize the quantum coherence and entanglement of the system. We show that this finding is not limited to disordered or high temperature systems but is instead a fundamental consequence of spontaneous parity time-reversal symmetry breaking associated with the quantum-to-classical transition. We then examine the effects of vibrational fluctuations, revealing a strong dephasing assisted transport enhancement for delocalized excitations in the presence of cooperative interactions. Our results highlight the rich, emergent behavior associated with decoherence and may be relevant to the study of biological photosynthetic antenna complexes or to the design of room-temperature quantum devices., Comment: Updated acknowledgements

In real-world machine learning (ML) pipelines, datasets are continuously growing. Models must incorporate this new training data to improve generalization and adapt to potential distribution shifts. The cost of model retraining is proportional to how frequently the model is retrained and how much data it is trained on, which makes the naive approach of retraining from scratch each time impractical. We present Modyn, a data-centric end-to-end machine learning platform. Modyn's ML pipeline abstraction enables users to declaratively describe policies for continuously training a model on a growing dataset. Modyn pipelines allow users to apply data selection policies (to reduce the number of data points) and triggering policies (to reduce the number of trainings). Modyn executes and orchestrates these continuous ML training pipelines. The system is open-source and comes with an ecosystem of benchmark datasets, models, and tooling. We formally discuss how to measure the performance of ML pipelines by introducing the concept of composite models, enabling fair comparison of pipelines with different data selection and triggering policies. We empirically analyze how various data selection and triggering policies impact model accuracy, and also show that Modyn enables high throughput training with sample-level data selection., Comment: final version published at SIGMOD'25; 30 pages

Evaluating explanations of image classifiers regarding ground truth, e.g. segmentation masks defined by human perception, primarily evaluates the quality of the models under consideration rather than the explanation methods themselves. Driven by this observation, we propose a framework for $\textit{jointly}$ evaluating the robustness of safety-critical systems that $\textit{combine}$ a deep neural network with an explanation method. These are increasingly used in real-world applications like medical image analysis or robotics. We introduce a fine-tuning procedure to (mis)align model$\unicode{x2013}$explanation pipelines with ground truth and use it to quantify the potential discrepancy between worst and best-case scenarios of human alignment. Experiments across various model architectures and post-hoc local interpretation methods provide insights into the robustness of vision transformers and the overall vulnerability of such AI systems to potential adversarial attacks.

This article shows how to combine the relative entropy method by D. Bresch, P.-E. Jabin, and Z. Wang in arXiv:1706.09564, arXiv:1906.04093 and the regularized $L^2(\mathbb{R}^d)$-estimate by Oelschl\"ager (Probability theory and related fields, 1987) to prove a strong propagation of chaos result for the viscous porous medium equation from a moderately interacting particle system in $L^\infty(0,T; L^1(\mathbb{R}^d))$-norm. In the moderate interacting setting, the interacting potential is a smoothed Dirac Delta distribution, however, current results regarding the relative entropy methods for singular potentials do not apply. The result holds on $\mathbb{R}^d$ for any dimension $d\geq 1$ and provides a quantitative result where the rate of convergence depends on the moderate scaling parameter and the dimension $d\geq 1$. Additionally, the presented method can be adapted for moderately interacting systems for which a certain convergence probability holds -- thus a propagation of chaos result in relative entropy can be obtained for kernels approximating Coulomb potentials., Comment: 20 pages

As systems based on opaque Artificial Intelligence (AI) continue to flourish in diverse real-world applications, understanding these black box models has become paramount. In response, Explainable AI (XAI) has emerged as a field of research with practical and ethical benefits across various domains. This paper not only highlights the advancements in XAI and its application in real-world scenarios but also addresses the ongoing challenges within XAI, emphasizing the need for broader perspectives and collaborative efforts. We bring together experts from diverse fields to identify open problems, striving to synchronize research agendas and accelerate XAI in practical applications. By fostering collaborative discussion and interdisciplinary cooperation, we aim to propel XAI forward, contributing to its continued success. Our goal is to put forward a comprehensive proposal for advancing XAI. To achieve this goal, we present a manifesto of 27 open problems categorized into nine categories. These challenges encapsulate the complexities and nuances of XAI and offer a road map for future research. For each problem, we provide promising research directions in the hope of harnessing the collective intelligence of interested stakeholders.

Industrial manufacturing is currently amidst it's fourth great revolution, pushing towards the digital transformation of production processes. One key element of this transformation is the formalization and digitization of processes, creating an increased potential to monitor, understand and optimize existing processes. However, one major obstacle in this process is the increased diversification and specialisation, resulting in the dependency on multiple experts, which are rarely amalgamated in small to medium sized companies. To mitigate this issue, this paper presents a novel approach for multi-criteria optimization of workflow-based assembly tasks in manufacturing by combining a workflow modeling framework and the HeuristicLab optimization framework. For this endeavour, a new generic problem definition is implemented in HeuristicLab, enabling the optimization of arbitrary workflows represented with the modeling framework. The resulting Pareto front of the multi-criteria optimization provides the decision makers a set of optimal workflows from which they can choose to optimally fit the current demands. The advantages of the herein presented approach are highlighted with a real world use case from an ongoing research project., Comment: This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this contribution is published in Computer Aided Systems Theory - EUROCAST 2022 and is available online at this URL https://doi.org/10.1007/978-3-031-25312-6_5

Efficient transport and harvesting of excitation energy under low light conditions is an important process in nature and quantum technologies alike. Here we formulate a quantum optics perspective to excitation energy transport in configurations of two-level quantum emitters with a particular emphasis on efficiency and robustness against disorder. We study a periodic geometry of emitter rings with subwavelength spacing, where collective electronic states emerge due to near-field dipole-dipole interactions. The system gives rise to collective subradiant states that are particularly suited to excitation transport and are protected from energy disorder and radiative decoherence. Comparing ring geometries with other configurations shows that that the former are more efficient in absorbing, transporting, and trapping incident light. Because our findings are agnostic as to the specific choice of quantum emitters, they indicate general design principles for quantum technologies with superior photon transport properties and may elucidate potential mechanisms resulting in the highly efficient energy transport efficiencies in natural light-harvesting systems., Comment: 12 pages, 6 figures

An increasingly salient public policy challenge is how to manage the growing number of satellites in orbit, including large constellations. Many policy initiatives have been proposed that attempt to address the problem from different angles, but there is a paucity of analytical tools to help policymakers evaluate the efficacy of these different proposals and any potential counterproductive outcomes. To help address this problem, this paper summarizes work done to develop an experimental integrated assessment model -- Orbital Debris Propagators Unified with Economic Systems (OPUS) -- that combines both astrodynamics of the orbital population and economic behavior of space actors. For a given set of parameters, the model first utilizes a given astrodynamic propagator to assess the state of objects in orbit. It then uses a set of user-defined economic and policy parameters -- e.g. launch prices, disposal regulations -- to model how actors will respond to the economic incentives created by a given scenario. For the purposes of testing, the MIT Orbital Capacity Tool (MOCAT) version 4S was used as the primary astrodynamics propagator to simulate the true expected target collision probability ($p_c$) for a given end-of-life (EOL) disposal plan. To demonstrate propagator-agnosticism, a Gaussian mixture probability hypothesis density (GMPHD) filter was also used to simulate $p_c$. We also explore economic policy instruments to improve both sustainability of and economic welfare from orbit use. In doing so, we demonstrate that this hybrid approach can serve as a useful tool for evaluating policy proposals for managing orbital congestion. We also discuss areas where this work can be made more robust and expanded to include additional policy considerations., Comment: This work was supported by 2022 grant funding from the NASA ROSES program. Code and simulation data available at https://github.com/akhilrao/OPUS

Explainable AI (XAI) is an increasingly important area of machine learning research, which aims to make black-box models transparent and interpretable. In this paper, we propose a novel approach to XAI that uses the so-called counterfactual paths generated by conditional permutations of features. The algorithm measures feature importance by identifying sequential permutations of features that most influence changes in model predictions. It is particularly suitable for generating explanations based on counterfactual paths in knowledge graphs incorporating domain knowledge. Counterfactual paths introduce an additional graph dimension to current XAI methods in both explaining and visualizing black-box models. Experiments with synthetic and medical data demonstrate the practical applicability of our approach.

Large language models, e.g. ChatGPT are currently contributing enormously to make artificial intelligence even more popular, especially among the general population. However, such chatbot models were developed as tools to support natural language communication between humans. Problematically, it is very much a ``statistical correlation machine" (correlation instead of causality) and there are indeed ethical concerns associated with the use of AI language models such as ChatGPT, such as Bias, Privacy, and Abuse. This paper highlights specific ethical concerns on ChatGPT and articulates key challenges when ChatGPT is used in various applications. Practical commandments for different stakeholders of ChatGPT are also proposed that can serve as checklist guidelines for those applying ChatGPT in their applications. These commandment examples are expected to motivate the ethical use of ChatGPT., Comment: 8 pages, 2 figures

We study the photophysics of molecular aggregates from a quantum optics perspective, with emphasis on deriving scaling laws for the fast non-radiative relaxation of collective electronic excitations, referred to as Kasha's rule. Aggregates exhibit an energetically broad manifold of collective states with delocalized electronic excitations originating from near field dipole-dipole exchanges between neighboring monomers. Photo-excitation at optical wavelengths, much larger than the monomer-monomer average separation, addresses almost exclusively symmetric collective states, which for an arrangement known as H-aggregate, show an upward hypsochromic shift. The extremely fast subsequent non-radiative relaxation via intramolecular vibrational modes populates lower energy, subradiant states, resulting in an effective inhibition of fluorescence. Our analytical treatment allows for the derivation of an approximate scaling law of this relaxation process, linear in the number of available low energy vibrational modes and directly proportional to the dipole-dipole interaction strength between neighbouring monomers.

Theoretical background: early verbal development is not yet fully understood, especially in its formative phase. Research question: can a reliable, easy-to-use coding scheme for the classification of early infant vocalizations be defined that is applicable as a basis for further analysis of language development? Methods: in a longitudinal study of 45 neurotypical infants, we analyzed vocalizations of the first 4 months of life. Audio segments were assigned to 5 classes: (1) Voiced and (2) Voiceless vocalizations; (3) Defined signal; (4) Non-target; (5) Nonassignable. Results: Two female coders with different experience achieved high agreement without intensive training. Discussion and Conclusion: The reliable scheme can be used in research and clinical settings for efficient coding of infant vocalizations, as a basis for detailed manual and machine analyses., Comment: This paper is in German