Search

We prove a Lipschitz-volume rigidity result for $1$-Lipschitz maps of non-zero degree between metric manifolds (metric spaces homeomorphic to a closed oriented manifold) and Riemannian manifolds. The proof is based on degree theory and recent developments of Lipschitz-volume rigidity for integral currents.

Over the past decade, autonomous stabilization of bosonic qubits has emerged as a promising approach for hardware-efficient protection of quantum information. However, applying these techniques to more complex encodings than the Schr\"odinger cat code requires exquisite control of high-order wave mixing processes. The challenge is to enable specific multiphotonic dissipation channels while avoiding unintended non-linear interactions. In this work, we leverage a genuine six-wave mixing process enabled by a near Kerr-free Josephson element to enforce dissipation of quartets of excitations in a high-impedance superconducting resonator. Owing to residual non-linearities stemming from stray inductances in our circuit, this dissipation channel is only effective when the resonator holds a specific number of photons. Applying it to the fourth excited state of the resonator, we show an order of magnitude enhancement of the state decay rate while only marginally impacting the relaxation and coherence of lower energy states. Given that stray inductances could be strongly reduced through simple modifications in circuit design and that our methods can be adapted to activate even higher-order dissipation channels, these results pave the way toward the dynamical stabilization of four-component Schr\"odinger cat qubits and even more complex bosonic qubits.

The Bose-Hubbard model effectively describes bosons on a lattice with on-site interactions and nearest-neighbour hopping, serving as a foundational framework for understanding strong particle interactions and the superfluid to Mott insulator transition. This paper aims to rigorously establish the validity of a mean-field approximation for the dynamics of quantum systems in high dimension, using the Bose--Hubbard model on a square lattice as a case study. We prove a trace norm estimate between the one-lattice-site reduced density of the Schr\"odinger dynamics and the mean-field dynamics in the limit of large dimension. Here, the mean-field approximation is in the hopping amplitude and not in the interaction, leading to a very rich and non-trivial mean-field equation. This mean-field equation does not only describe the condensate, as is the case when the mean-field description comes from a large particle number limit averaging out the interaction, but it allows for a phase transition to a Mott insulator since it contains the full non-trivial interaction. Our work is a rigorous justification of a simple case of the highly successful dynamical mean-field theory (DMFT) for bosons, which somewhat surprisingly yields many qualitatively correct results in three dimensions.

Recently-synthesised MoSi$_2$N$_4$ is the first septuple layer two-dimensional material, which doesn't naturally occurs as a layered crystal, and has been obtained with CVD growth. It can be considered as MoN$_2$ crystal (with a crystal structure of MoS2) intercalating Si2N2 two-dimensional layer (with the structure similar to InSe). Such classification gave rise to the understanding of the electronic properties of the material, but also to the prediction of other members of the family (many dozens of them) as well as to the way to classify those. Whereas the originally-synthesised MoSi$_2$N$_4$ is a semiconductor, some of the members of the family are also metallic and some even demonstrate magnetic properties. Interestingly, the room-temperature mobility predicted for such crystals can be as high few thousands cm$^2$/V*s (hole mobility typically higher than electron) with some record cases as high as $5\times 10^4$ cm$^2$/V*s, making these materials strong contenders for future electronic applications. The major interest towards these materials is coming from the septuple layer structure, which allows multiple crystal phases, but also complex compositions, in particular those with broken mirror-reflection symmetry against the layer of metal atoms.

This paper presents a novel approach for the automatic generation of Cued Speech (ACSG), a visual communication system used by people with hearing impairment to better elicit the spoken language. We explore transfer learning strategies by leveraging a pre-trained audiovisual autoregressive text-to-speech model (AVTacotron2). This model is reprogrammed to infer Cued Speech (CS) hand and lip movements from text input. Experiments are conducted on two publicly available datasets, including one recorded specifically for this study. Performance is assessed using an automatic CS recognition system. With a decoding accuracy at the phonetic level reaching approximately 77%, the results demonstrate the effectiveness of our approach.

Let $S(n)$ be a real valued random walk with i.i.d. increments which have zero mean and finite variance. We are interested in the asymptotic properties of the stopping time $T(g):=\inf\{n\ge1: S(n)\le g(n)\}$, where $g(t)$ is a boundary function. In the present paper we deal with the parametric family of boundaries $\{g_{a,b}(t)=c\sqrt{t+b}-a, b\ge0, a>c\sqrt{b}\}$. First, assuming that sufficiently many moments of increments of the walk are finite, we construct a positive space-time harmonic function $W(a,b)$. Then we show that there exist $p(c)>0$ and a constant $\varkappa(c)$ such that $\mathbf{P}(T_{g_{a,b}}>n)\sim \varkappa(c)\frac{W(a,b)}{n^{p(c)/2}}$ as $n\to\infty$., Comment: 33 pages

The abundance of information sources in our digital environment makes it difficult to study how such information shapes individuals' attitudes towards current policies. Our study investigates how self-directed online search in a naturalistic setting through three randomized controlled experiments with 791 German participants on three topical policy issues: basic child support, renewable energy transition, and cannabis legalization. Participants' online browsing was passively tracked, and changes in their attitudes were measured. By encouraging participants to seek online information, this study enhances ecological validity compared to traditional experiments that expose subjects to predetermined content. Significant attitude shifts were observed for child support and cannabis legalization, but not for renewable energy transition. Some findings suggest that the specificity and granularity of policy topics may affect whether and how online information shapes political views, providing insights into the nuanced impact of online information seeking on policy attitudes. By exploring participant's searches and visits, we depict the behavioral patterns that emerge on from our encouragement. Our experimental approach lays the groundwork for future research to advance understanding of the media effect within the dynamic online information landscape.

Electrons trapped on the surface of superfluid helium have been thought of as a useful resource for quantum computing. Such electrons show long coherence of their surface-bound (Rydberg) states combined with their easy electrostatic manipulation. Recent proposals explored the possibility of coupling the spin state of an electron and the state of its quantized motion with a promise of a highly scalable 2D architecture for a quantum computer. However, despite recent progress in the detection of quantized lateral motion of electrons using a circuit-QED setup, the manipulation of a small number of electrons and their quantum state detection remains a challenging problem. Here, we report on the detection of the Rydberg transition of electrons on superfluid helium in an on-chip microchannel device in which electrons are moved and trapped by a set of electrostatic gates. A highly sensitive image-charge detection system allows us not only to resolve the transition spectra of such electrons, but also to perform the device characterization. The demonstrated sensitivity shows the feasibility of detecting the Rydberg transition of a single electron, which can open a new pathway for a non-destructive spin-state readout., Comment: 12 pages, 13 figures

We report the effect of nonlinear bias of the frequency of collective oscillations of sin-coupled phase oscillators subject to individual asymmetric Cauchy noises. The noise asymmetry makes the Ott-Antonsen Ansatz inapplicable. We argue that, for all stable non-Gaussian noises, the tail asymmetry is not only possible (in addition to the trivial shift of the distribution median) but also generic in many physical and biophysical set-ups. For the theoretical description of the effect, we develop a mathematical formalism based on the circular cumulants. The derivation of rigorous asymptotic results can be performed on this basis but seems infeasible in traditional terms of the circular moments (the Kuramoto-Daido order parameters). The effect of the entrainment of individual oscillator frequencies by the global oscillations is also reported in detail. The accuracy of theoretical results based on the low dimensional circular cumulant reductions is validated with the high-accuracy "exact" solutions calculated with the continued fraction method., Comment: 14 pages, 6 figures

Complex geometric tasks such as geometric modeling, physical simulation, and texture parametrization often involve the embedding of many complex sub-domains with potentially different dimensions. These tasks often require evolving the geometry and topology of the discretizations of these sub-domains, and guaranteeing a \emph{consistent} overall embedding for the multiplicity of sub-domains is required to define boundary conditions. We propose a data structure and algorithmic framework for hierarchically encoding a collection of meshes, enabling topological and geometric changes to be automatically propagated with coherent correspondences between them. We demonstrate the effectiveness of our approach in surface mesh decimation while preserving UV seams, periodic 2D/3D meshing, and extending the TetWild algorithm to ensure topology preservation of the embedded structures.

We report on the electron spin resonance (ESR), heat capacity, magnetization, nuclear magnetic resonance (NMR), magnetic circular and linear dichroism (XMCD, XMLD), as well as the electrical resistivity of EuMn$_{2}$P$_{2}$ single crystals. Antiferromagnetic order of Eu was observed in several quantities at $T^{\rm Eu}_{\rm N}\,=\,18\,\rm K$. The temperature dependencies of ESR linewidth and resonance shift show, when approaching the Eu-ordered state, a divergence towards $T^{\rm Eu}_{\rm N}$, indicating the growing importance of magnetic correlations and the build-up of internal magnetic fields. An additional temperature scale of $\approx 47\,\rm K$ has considerable impact on linewidth, resonance field and intensity. This points to the presence of weak Mn-based ordering. The observed ESR line is interpreted as an Eu$^{2+}$ resonance, which probes the weak magnetic background of the Mn subsystem. Such picture is suggested by the lineshape which keeps to be Lorentzian across the $47\,\rm K$ scale and by the ESR intensity which can be described by the same Curie-Weiss temperature above and below $47\,\rm K$. In the same temperature range anomalies were observed at $48.5\,\rm K$ and $51\,\rm K$ in the heat capacity data as well as a pronounced broadening of the NMR signal of the EuMn$_{2}$P$_{2}$ samples. In XMCD and XMLD measurements, this weak magnetic order could not be detected in the same temperature range which might be due to the small magnetic moment, with a potential $c$-component or frustration.

Fundamental weak interactions have been shown to violate parity in both nuclear and atomic systems. However, observation of parity violation in a molecular system has proven an elusive target. Nuclear spin dependent contributions of the weak interaction are expected to result in energetic differences between enantiomers manifesting in nuclear magnetic resonance (NMR) spectra as chemical shift differences on the order of $10^{-6}$ Hz to $10^{-3}$ Hz for high-$Z$ nuclei. By employing simultaneous measurements of the diastereomeric splittings for a light and a heavy nucleus in solution-state NMR, residual chemical shift differences persisting in non-chiral environment between enantiomers of chiral compounds smaller than the typical linewidth of high-field NMR may be resolved. Sources of error must be identified and minimized to verify that the observed effect is, in fact, due to parity violation and not systematic effects. This paper presents a detailed analysis of a system incorporating \textsuperscript{31}P and \textsuperscript{1}H NMR to elucidate the systematic effects and to guide experiments with higher-$Z$ nuclei where molecular parity violation may be resolved.

Analytically solving the magnetostatic Maxwell equations in the bispherical coordinates, we calculate the magnetic field around two uniformly magnetized spheres oriented so that their magnetic moments are parallel to the axis passing through the centers of the spheres. We demonstrate that, contrary to what is often claimed in the literature, the magnetic interaction between such spheres is not equivalent to the interaction between two point magnetic dipoles placed in the centers of the spheres. The nonzero levitation force acting on a uniformly magnetized sphere or a point magnetic dipole above a superconducting sphere in the ideal Meissner state is a clear manifestation of the non-equivalence., Comment: 22 pages, 2 figures

Optimization-based (OB) alternatives to traditional flux limiters couch preservation of properties such as local bounds and maximum principles into optimization problems, which impose these properties through inequality constraints. In this paper, we propose a new potential-target OB approach that enforces these properties using an optimal control formulation, in which the control is the source term expressed through flux potentials. The resulting OB formulation combines superb accuracy with excellent local conservation properties, but complicates the development of scalable iterative solvers, which is greatly influenced by the choice of semi-norms for the objective function. We use this fact to design scalable iterative solvers based on matrix-free trust-region Newton methods with projections onto convex sets. These solvers leverage inexpensive multigrid V-cycles while satisfying all constraints to machine precision. Numerical experiments reveal that the convergence behavior of the solvers can be greatly improved by a simple scaling of the inequality constraints. We demonstrate excellent performance in applications to linear test problems, such as $L^2$ projection and solid body rotation, and to the Cahn-Hilliard equation.

SPARC4 is a new astronomical instrument developed entirely by Brazilian institutions, currently installed on the 1.6-m Perkin-Elmer telescope of the Pico dos Dias Observatory. It allows the user to perform photometric or polarimetric observations simultaneously in the four SDSS bands (g, r, i, and z). In this paper, we describe the control system developed for SPARC4. This system is composed of S4ACS, S4ICS, and S4GUI softwares and associated hardware. S4ACS is responsible for controlling the four EMCCD scientific cameras (one for each instrument band). S4ICS controls the sensors and motors responsible for the moving parts of SPARC4. Finally, S4GUI is the interface used to perform observations, which includes the choice of instrument configuration and image acquisition parameters. S4GUI communicates with the instrument subsystems and with some observatory facilities, needed during the observations. Bench tests were performed for the determination of the overheads added by SPARC4 control system in the acquisition of photometric and polarimetric series of images. In the photometric mode, SPARC4 allows the acquisition of a series of 1400 full-frame images, with a deadtime of 4.5 ms between images. Besides, several image series can be concatenated with a deadtime of 450 ms plus the readout time of the last image. For the polarimetric mode, measurements can be obtained with a deadtime of 1.41 s plus the image readout time between subsequent waveplate positions. For both photometric and polarimetric modes, the user can choose among operating modes with image readout times between 5.9 ms and 1.24 s, which ultimately defines the instrument temporal performance., Comment: 20 pages, 14 figures, peer-reviewed paper

Nonlinear differential equations and systems play a crucial role in modeling systems where time-dependent factors exhibit nonlinear characteristics. Due to their nonlinear nature, solving such systems often presents significant difficulties and challenges. In this study, we propose a method utilizing Physics-Informed Neural Networks (PINNs) to solve the nonlinear energy supply-demand (ESD) system. We design a neural network with four outputs, where each output approximates a function that corresponds to one of the unknown functions in the nonlinear system of differential equations describing the four-dimensional ESD problem. The neural network model is then trained and the parameters are identified, optimized to achieve a more accurate solution. The solutions obtained from the neural network for this problem are equivalent when we compare and evaluate them against the Runge-Kutta numerical method of order 4/5 (RK45). However, the method utilizing neural networks is considered a modern and promising approach, as it effectively exploits the superior computational power of advanced computer systems, especially in solving complex problems. Another advantage is that the neural network model, after being trained, can solve the nonlinear system of differential equations across a continuous domain. In other words, neural networks are not only trained to approximate the solution functions for the nonlinear ESD system but can also represent the complex dynamic relationships between the system's components. However, this approach requires significant time and computational power due to the need for model training., Comment: Submitted to Computation J

As deep learning models become larger and more expensive, many practitioners turn to fine-tuning APIs. These web services allow fine-tuning a model between two parties: the client that provides the data, and the server that hosts the model. While convenient, these APIs raise a new concern: the data of the client is at risk of privacy breach during the training procedure. This challenge presents an important practical case of vertical federated learning, where the two parties perform parameter-efficient fine-tuning (PEFT) of a large model. In this study, we systematically search for a way to fine-tune models over an API while keeping the labels private. We analyze the privacy of LoRA, a popular approach for parameter-efficient fine-tuning when training over an API. Using this analysis, we propose P$^3$EFT, a multi-party split learning algorithm that takes advantage of existing PEFT properties to maintain privacy at a lower performance overhead. To validate our algorithm, we fine-tune DeBERTa-v2-XXLarge, Flan-T5 Large and LLaMA-2 7B using LoRA adapters on a range of NLP tasks. We find that P$^3$EFT is competitive with existing privacy-preserving methods in multi-party and two-party setups while having higher accuracy.

Excitons in Transition Metal Dichalcogenides (TMDs) acquire a spin-like quantum number, a pseudospin, originating from the crystal's discrete rotational symmetry. Here, we break this symmetry using a tunable uniaxial strain, effectively generating a pseudomagnetic field exceeding 40 Tesla. Under this large field, we demonstrate pseudospin analogs of spintronic phenomena such as the Zeeman effect and Larmor precession. Moreover, we determine previously inaccessible fundamental properties of TMDs, including the strength of the depolarizing field responsible for the loss of exciton coherence. Finally, we uncover the bosonic -- as opposed to fermionic -- nature of many-body excitonic species using the pseudomagnetic equivalent of the $g$-factor spectroscopy. Our work is the first step toward establishing this spectroscopy as a universal method for probing correlated many-body states and realizing pseudospin analogs of spintronic devices.

We consider three conditions on metric manifolds with finite volume: (1) the existence of a metric fundamental class, (2) local index bounds for Lipschitz maps, and (3) Gromov--Hausdorff approximation with volume control by bi-Lipschitz manifolds. Condition (1) is known for metric manifolds satisfying the LLC condition by work of Basso--Marti--Wenger, while (3) is known for metric surfaces by work of Ntalampekos--Romney. We prove that for metric manifolds with finite Nagata dimension, all three conditions are equivalent and that without assuming finite Nagata dimension, (1) implies (2) and (3) implies (1). As a corollary we obtain a generalization of the approximation result of Ntalampekos--Romney to metric manifolds of dimension $n\ge 2$, which have the LLC property and finite Nagata dimension., Comment: 25 pages, comments welcome!

Dynamic fetal heart magnetic resonance imaging (MRI) presents unique challenges due to the fast heart rate of the fetus compared to adult subjects and uncontrolled fetal motion. This requires high temporal and spatial resolutions over a large field of view, in order to encompass surrounding maternal anatomy. In this work, we introduce Dynamic Cardiac Reconstruction Attention Network (DCRA-Net) - a novel deep learning model that employs attention mechanisms in spatial and temporal domains and temporal frequency representation of data to reconstruct the dynamics of the fetal heart from highly accelerated free-running (non-gated) MRI acquisitions. DCRA-Net was trained on retrospectively undersampled complex-valued cardiac MRIs from 42 fetal subjects and separately from 153 adult subjects, and evaluated on data from 14 fetal and 39 adult subjects respectively. Its performance was compared to L+S and k-GIN methods in both fetal and adult cases for an undersampling factor of 8x. The proposed network performed better than the comparators for both fetal and adult data, for both regular lattice and centrally weighted random undersampling. Aliased signals due to the undersampling were comprehensively resolved, and both the spatial details of the heart and its temporal dynamics were recovered with high fidelity. The highest performance was achieved when using lattice undersampling, data consistency and temporal frequency representation, yielding PSNR of 38 for fetal and 35 for adult cases. Our method is publicly available at https://github.com/denproc/DCRA-Net.

A parameter $\psi$ was recently defined and introduced into the epidemiological modelling scientific literature, and is being accepted. The said parameter was used to argue that there was a disproportionate risk of infection incurred by vaccinated persons due to contacts with unvaccinated persons during the declared COVID-19 pandemic. Opposing published results show that, in general, there is virtually never a disproportionate risk to the vaccinated from the unvaccinated during a respiratory pandemic. Here, we show that the newly introduced vaccinology parameter $\psi$ is incorrectly defined and that the conclusions of disproportionate risk are not valid. Specifically, we prove that the originating authors Fisman et al. (2022, 2024) incorrectly defined and applied the parameter $\psi$. Their application would imply that the said risk increases with increasing segregation from the unvaccinated (up to complete segregation), increases with increasing vaccination coverage (up to complete coverage) and increases with increasing vaccine efficacy (up to perfect vaccine efficacy), which is impossible. Use of the erroneous parameter $\psi$ has the potential to encourage unnecessarily aggressive public health policies and interventions., Comment: 23 pages, 3 figures, 1 table, 11 references

Balancing high sensitivity with a broad dynamic range is a fundamental challenge in measurement science, as improving one often compromises the other. While traditional quantum metrology has prioritized enhancing local sensitivity, a large dynamic range is crucial for applications such as atomic clocks, where extended phase interrogation times contribute to wider phase range. In this Letter, we introduce a novel quantum deamplification mechanism that extends dynamic range at a minimal cost of sensitivity. Our approach uses two sequential spin-squeezing operations to generate and detect an entangled probe state, respectively. We demonstrate that the optimal quantum interferometer limit can be approached through two-axis counter-twisting dynamics. Further expansion of dynamic range is possible by using sequential quantum deamplification interspersed with phase encoding processes. Additionally, we show that robustness against detection noise can be enhanced by a hybrid sensing scheme that combines quantum deamplification with quantum amplification. Our protocol is within the reach of state-of-the-art atomic-molecular-optical platforms, offering a scalable, noise-resilient pathway for entanglement-enhanced metrology., Comment: (4.5+2.5) pages, 4 figures

In large-scale software systems, there are often no fully-fledged bug reports with human-written descriptions when an error occurs. In this case, developers rely on stack traces, i.e., series of function calls that led to the error. Since there can be tens and hundreds of thousands of them describing the same issue from different users, automatic deduplication into categories is necessary to allow for processing. Recent works have proposed powerful deep learning-based approaches for this, but they are evaluated and compared in isolation from real-life workflows, and it is not clear whether they will actually work well at scale. To overcome this gap, this work presents three main contributions: a novel model, an industry-based dataset, and a multi-faceted evaluation. Our model consists of two parts - (1) an embedding model with byte-pair encoding and approximate nearest neighbor search to quickly find the most relevant stack traces to the incoming one, and (2) a reranker that re-ranks the most fitting stack traces, taking into account the repeated frames between them. To complement the existing datasets collected from open-source projects, we share with the community SlowOps - a dataset of stack traces from IntelliJ-based products developed by JetBrains, which has an order of magnitude more stack traces per category. Finally, we carry out an evaluation that strives to be realistic: measuring not only the accuracy of categorization, but also the operation time and the ability to create new categories. The evaluation shows that our model strikes a good balance - it outperforms other models on both open-source datasets and SlowOps, while also being faster on time than most. We release all of our code and data, and hope that our work can pave the way to further practice-oriented research in the area., Comment: Published at SANER'25. 11 pages, 2 figures

Transformer-based approaches such as BERT4Rec and SASRec demonstrate strong performance in Next Item Recommendation (NIR) tasks. However, applying these architectures to Next-Basket Recommendation (NBR) tasks, which often involve highly repetitive interactions, is challenging due to the vast number of possible item combinations in a basket. Moreover, frequency-based methods such as TIFU-KNN and UP-CF still demonstrate strong performance in NBR tasks, frequently outperforming deep-learning approaches. This paper introduces SAFERec, a novel algorithm for NBR that enhances transformer-based architectures from NIR by incorporating item frequency information, consequently improving their applicability to NBR tasks. Extensive experiments on multiple datasets show that SAFERec outperforms all other baselines, specifically achieving an 8\% improvement in Recall@10.

We introduce two topological non-$\Sigma$ operad structures on planar line arrangements subject to a certain geometric order condition, ensuring a well-defined notion of particle ordering on a distinguished line. This is interpreted in terms of scattering diagrams in purely elastic (1+1)-dimensional theories. We discuss a possible approach to factorized scattering in operadic terms.

We report a novel nanoscale device concept based on a highly doped $\delta$-layer tunnel junction embedded in a semiconductor for charge sensing. Recent advances in Atomic Precision Advanced Manufacturing (APAM) processes have enabled the fabrication of devices based on quasi-2D, highly conductive, highly doped regions in semiconductor materials. In this work, we demonstrate that FET-based sensors utilizing APAM $\delta$-layer tunnel junctions are ultrasensitive to the presence of charges near the tunnel junction, allowing the use of these devices for detecting charges by observing changes in the electrical current. We demonstrate that these devices can enhance the sensitivity in the limit, i.e. for small concentrations of charges, exhibiting significantly superior sensitivity compared to traditional FET-based sensors. We also propose that the extreme sensitivity arises from the strong quantization of the conduction band in these highly-confined systems.

View materialization, index selection, and plan caching are well-known techniques for optimization of query processing in database systems. The essence of these tasks is to select and save a subset of the most useful candidates (views/indexes/plans) for reuse within given space/time budget constraints. In this paper, based on the View Selection Problem, we propose a unified view on these problems. We identify the root causes of the complexity of these selection problems and provide a detailed analysis of techniques to cope with them. Our survey provides a modern classification of selection algorithms known in the literature, including the latest ones based on Machine Learning. We provide a ground for the reuse of the selection techniques between different optimization scenarios and highlight challenges and promising directions in the field.

The capability to tailor mutual interactions between colloidal nanoparticles strongly depends on the length scales involved. While electrostatic and optomechanically driven interactions can cover nano and micron-scale landscapes, controlling inter-particle dynamics at larger distances remains a challenge. Small physical and electromagnetic cross-sections of nanoparticles make long-range interactions, screened by a fluid environment, inefficient. To bypass the limitations, we demonstrate that forming micron-scale bubbles around gilded vaterite nanoparticles enables mediating long-range interactions via thermo-optical forces. Femtosecond laser illumination leads to the encapsulation of light-absorbing particles inside long-lasting micron-scale bubbles, which in turn behave as negative lenses refracting incident light. Our experiments reveal the bubble-induced collimation of laser beams, traversing over mm-scale distances. The collimated beams are visualized with the aid of phase-contrast Schlieren imaging, which reveals refractive index variations, caused by temperature gradients within the fluid. We demonstrate that the refracted beams initiate the formation of secondary bubbles around nearby gilded vaterite particles. As the consequence, we demonstrate the ability to control secondary bubble motion by pushing and pulling it with optical radiation pressure force and by thermocapillary Marangoni effect, respectively. The latter facilitates interactions over millimeter-scale distances, which are otherwise unachievable. Apart from exploring new physical effects, mediating long-range interactions can find a use in a range of applications including drug design and screening, photochemistry, design of colloidal suspensions, and many others., Comment: 6 figures, 12 pages + references

Tokenization is a necessary component within the current architecture of many language models, including the transformer-based large language models (LLMs) of Generative AI, yet its impact on the model's cognition is often overlooked. We argue that LLMs demonstrate that the Distributional Hypothesis (DH) is sufficient for reasonably human-like language performance, and that the emergence of human-meaningful linguistic units among tokens motivates linguistically-informed interventions in existing, linguistically-agnostic tokenization techniques, particularly with respect to their roles as (1) semantic primitives and as (2) vehicles for conveying salient distributional patterns from human language to the model. We explore tokenizations from a BPE tokenizer; extant model vocabularies obtained from Hugging Face and tiktoken; and the information in exemplar token vectors as they move through the layers of a RoBERTa (large) model. Besides creating sub-optimal semantic building blocks and obscuring the model's access to the necessary distributional patterns, we describe how tokenization pretraining can be a backdoor for bias and other unwanted content, which current alignment practices may not remediate. Additionally, we relay evidence that the tokenization algorithm's objective function impacts the LLM's cognition, despite being meaningfully insulated from the main system intelligence.

We consider a one-dimensional random walk $S_n$ having i.i.d. increments with zero mean and finite variance. We continue our study of asymptotic expansions for local probabilities $\mathbf P(S_n=x,\tau_0>n)$, which has been started in \cite{DTW23}. Obtained there expansions make sense in the zone $x=o(\frac{\sqrt{n}}{\log^{1/2} n})$ only. In the present paper we derive an alternative expansion, which deals with $x$ of order $\sqrt{n}$., Comment: 40 pages

We consider random walks conditioned to stay positive. When the mean of increments is zero and variance is finite it is known that they converge to the Rayleigh distribution. In the present paper we derive a Berry-Esseen type estimate and show that the rate of convergence is of order $n^{-1/2}$., Comment: 22 pages

Video diffusion models have achieved impressive realism and controllability but are limited by high computational demands, restricting their use on mobile devices. This paper introduces the first mobile-optimized video diffusion model. Starting from a spatio-temporal UNet from Stable Video Diffusion (SVD), we reduce memory and computational cost by reducing the frame resolution, incorporating multi-scale temporal representations, and introducing two novel pruning schema to reduce the number of channels and temporal blocks. Furthermore, we employ adversarial finetuning to reduce the denoising to a single step. Our model, coined as MobileVD, is 523x more efficient (1817.2 vs. 4.34 TFLOPs) with a slight quality drop (FVD 149 vs. 171), generating latents for a 14x512x256 px clip in 1.7 seconds on a Xiaomi-14 Pro. Our results are available at https://qualcomm-ai-research.github.io/mobile-video-diffusion/

Quantitative acoustic microscopy (QAM) is a cutting-edge imaging modality that leverages very high-frequency ultrasound to characterize the acoustic and mechanical properties of biological tissues at microscopic resolutions. Radio-frequency echo signals are digitized and processed to yield two-dimensional maps. This paper introduces a weighted Hankel-based spectral method with a reweighting strategy to enhance robustness with regard to noise and reduce unreliable acoustic parameter estimates. Additionally, we derive, for the first time in QAM, Cram\'er-Rao bounds to establish theoretical performance benchmarks for acoustic parameter estimation. Simulations and experimental results demonstrate that the proposed method consistently outperform standard autoregressive approach, particularly under challenging conditions. These advancements promise to improve the accuracy and reliability of tissue characterization, enhancing the potential of QAM for biomedical applications.

A programming library was developed, based on Stratton-Chu diffraction integrals for calculating reflected optical fields. Dipole-type focusing schemes with tunable number of beams and mirror placements were studied, considering the influence of phase distortion and aberrations. The intensity above $3\times 10^{26}$ W/cm$^2$ was found theoretically attainable in a system of 12 beams of 50 PW each with about 90 % of that value realistically achievable., Comment: 17 pages, 14 figures, Accepted for publication in Applied Optics

Superpositions of handed molecular states give rise to achiral eigenstates, delocalized across a double-well potential via tunneling. A coherent superposition of these energy eigenstates could dynamically relocalize the molecules into chiral states, which has only been addressed theoretically. Here, we present a microwave six-wave mixing pump-probe study to create and probe coherent chiral tunneling dynamics in a rotational state. Through a time-resolved scheme, we uncover the periodic time evolution of the induced chiral wavepacket under field-free conditions. Moreover, we demonstrate precise phase control of this coherence via phase modulation during pump excitation., Comment: 14 pages, 4 figures in the main text and 27 pages, 12 figures, and 4 tables in the Supplementary Material

A model of sound propagation in a magnetized magnetic fluid containing ellipsoidal aggregates is proposed. The model quantitatively describes the geometry of the aggregates formed from nanoparticles. Expressions for the attenuation coefficient and the sound velocity have been derived, taking into account dipole-dipole interaction between the aggregates. The model demonstrates good agreement with the experimental data. In limiting cases, the derived expressions reduce to classical ones, and the fitting parameters are merely geometric characteristics of the aggregates. The developed model enables the determination of aggregate sizes and the distances between them from an experiment that can be conducted without direct contact. The results obtained can also be used to analyze the field dependencies of acoustic properties and to model the kinetics of aggregate growth in a magnetic field., Comment: 10 pages, 5 figures

An effective approach for sampling from unnormalized densities is based on the idea of gradually transporting samples from an easy prior to the complicated target distribution. Two popular methods are (1) Sequential Monte Carlo (SMC), where the transport is performed through successive annealed densities via prescribed Markov chains and resampling steps, and (2) recently developed diffusion-based sampling methods, where a learned dynamical transport is used. Despite the common goal, both approaches have different, often complementary, advantages and drawbacks. The resampling steps in SMC allow focusing on promising regions of the space, often leading to robust performance. While the algorithm enjoys asymptotic guarantees, the lack of flexible, learnable transitions can lead to slow convergence. On the other hand, diffusion-based samplers are learned and can potentially better adapt themselves to the target at hand, yet often suffer from training instabilities. In this work, we present a principled framework for combining SMC with diffusion-based samplers by viewing both methods in continuous time and considering measures on path space. This culminates in the new Sequential Controlled Langevin Diffusion (SCLD) sampling method, which is able to utilize the benefits of both methods and reaches improved performance on multiple benchmark problems, in many cases using only 10% of the training budget of previous diffusion-based samplers.

Stable composite objects, such as hadrons, nuclei, atoms, molecules and superconducting pairs, formed by attractive forces are ubiquitous in nature. By contrast, composite objects stabilized by means of repulsive forces were long thought to be theoretical constructions owing to their fragility in naturally occurring systems. Surprisingly, the formation of bound atom pairs by strong repulsive interactions has been demonstrated experimentally in optical lattices. Despite this success, repulsively bound particle pairs were believed to have no analogue in condensed matter owing to strong decay channels. Here we present spectroscopic signatures of repulsively bound three-magnon states and bound magnon pairs in the Ising-like chain antiferromagnet BaCo$_2$V$_2$O$_8$. In large transverse fields, below the quantum critical point, we identify repulsively bound magnon states by comparing terahertz spectroscopy measurements to theoretical results for the Heisenberg-Ising chain antiferromagnet, a paradigmatic quantum many-body model. Our experimental results show that these high-energy repulsively bound magnon states are well separated from continua, exhibit significant dynamical responses and, despite dissipation, are sufficiently long-lived to be identified. As the transport properties in spin chains can be altered by magnon bound states, we envision such states could serve as resources for magnonics based quantum information processing technologies., Comment: 27 pages, 4 + 5 figures, accepted version

Electrically induced $p-n$ junctions in graphene bilayer (GBL) have shown superior performance for detection of sub-THz radiation at cryogenic temperatures, especially upon electrical induction of the band gap $E_g$. Still, the upper limits of responsivity and noise equivalent power (NEP) at very large $E_g$ remained unknown. Here, we study the cryogenic performance of GBL detectors at $f=0.13$ THz by inducing gaps up to $E_g \approx 90$ meV, a value close to the limits observed in recent transport experiments. High value of the gap is achieved by using high-$\kappa$ bottom hafnium dioxide gate dielectric. The voltage responsivity, current responsivity and NEP optimized with respect to doping do not demonstrate saturation with gap induction up to its maximum values. The NEP demonstrates an order-of-magnitude drop from $\sim450$ fW/Hz$^{1/2}$ in the gapless state to $\sim30$ fW/Hz$^{1/2}$ at the largest gap. At largest induced band gaps, plasmonic oscillations of responsivity become visible and important for optimization of sub-THz response.

Mid-infrared (mid-IR) photodetectors play a crucial role in various applications, including the development of biomimetic vision systems that emulate neuronal function. However, current mid-IR photodetector technologies are limited by their cost and efficiency. In this work, we demonstrate a new type of photodetector based on a tunnel structure made of two-dimensional materials. The effect manifests when the upper and lower layers of the tunnel structure are heated differently. The photoswitching is threshold-based and represents a ``jump'' in voltage to another branch of the current-voltage characteristic when illuminated at a given current. This mechanism provides enormous photovoltage (0.05$-$1~V) even under weak illumination. Our photodetector has built-in nonlinearity and is therefore an ideal candidate for use in infrared vision neurons. Additionally, using this structure, we demonstrated the possibility of selective heating of layers in a van der Waals stack using mid-IR illumination. This method will allow the study of heat transfer processes between layers of van der Waals structures, opening new avenues in the physics of phonon interactions.

Circuits based on sum-product structure have become a ubiquitous representation to compactly encode knowledge, from Boolean functions to probability distributions. By imposing constraints on the structure of such circuits, certain inference queries become tractable, such as model counting and most probable configuration. Recent works have explored analyzing probabilistic and causal inference queries as compositions of basic operators to derive tractability conditions. In this paper, we take an algebraic perspective for compositional inference, and show that a large class of queries - including marginal MAP, probabilistic answer set programming inference, and causal backdoor adjustment - correspond to a combination of basic operators over semirings: aggregation, product, and elementwise mapping. Using this framework, we uncover simple and general sufficient conditions for tractable composition of these operators, in terms of circuit properties (e.g., marginal determinism, compatibility) and conditions on the elementwise mappings. Applying our analysis, we derive novel tractability conditions for many such compositional queries. Our results unify tractability conditions for existing problems on circuits, while providing a blueprint for analysing novel compositional inference queries., Comment: NeurIPS 2024

Magnetic fluids exhibit tunable structures and electrophysical properties, making them promising for adaptive optical systems, biomedical sensors, and microelectromechanical devices. However, the dynamic evolution of their microstructure under varying magnetic fields remains insufficiently explored. This study investigates the structural and dielectric properties of transformer oil-based magnetic fluids containing 0.2-10 vol% magnetite nanoparticles, across a frequency range of 20 Hz to 10 MHz. Particular attention is given to the dynamics of aggregate reorientation in response to alternating magnetic fields. Experimental results demonstrate that low nanoparticle concentrations lead to a linear increase in dielectric permittivity and conductivity, consistent with the Maxwell-Wagner model. In contrast, higher concentrations exhibit conductivity saturation and dispersion effects due to the formation of elongated aggregates. An analysis based on the Boyle polarization model describes the relaxation and structural changes associated with aggregation dynamics. Changes in the magnetic field orientation induce aggregate reconfiguration and significant structural transformations. At early stages, elongated chains form, subsequently thickening until an equilibrium state is reached. Elevated temperatures accelerate these processes by reducing medium viscosity and aggregate order. The findings highlight the critical role of reorientation dynamics in designing high-speed magnetic sensors, vibration isolation systems, and adaptive devices operating in dynamic magnetic environments., Comment: 15 pages, 15 figures

For the Hamming graph $H(n,q)$, where a $q$ is a constant prime power and $n$ grows, we construct perfect colorings without non-essential arguments such that $n$ depends exponentially on the off-diagonal part of the quotient matrix. In particular, we construct unbalanced Boolean ($q=2$) functions such that the number of essential arguments depends exponentially on the degree of the function.

We investigate theoretically photonic chips with monolithically integrated blue light sources. According to our evaluations, a group-III nitride light emitting heterostructure can be efficiently combined with silicon nitride waveguiding layers. Low losses, high level of miniaturization and built-in light injection mechanism potentially make the selected platform attractive for applications and draw up an addition or even alternative to silicon photonics. We use a combination of drift-diffusion and Maxwell equations to build a model of the proposed multilayer structure. The model allows to choose the best parameters for both light emitting sandwich and waveguiding layers as well as to pick an optimal coupling regime. High transmittance coefficients are obtained. Various optimal geometries are analised with respect to captured power, desired polarization and excited modes. The obtained numbers are promising and allow in the future to pave the way towards hetero-integrated blue photonic circuits.

Spatial patterns are widely observed in numerous nonequilibrium natural systems, often undergoing complex transitions and bifurcations, thereby exhibiting significant importance in many physical and biological systems such as embryonic development, ecosystem desertification, and turbulence. However, how spatial pattern formation emerges and how the spatial pattern switches are not fully understood. Here, we developed a landscape-flux field theory via the spatial mode expansion method to uncover the underlying physical mechanism of the pattern formation and switching. We identified the landscape and flux field as the driving force for spatial dynamics and applied this theory to the critical transitions between spatial vegetation patterns in semi-arid ecosystems, revealing that the nonequilibrium flux drives the switchings of spatial patterns. We uncovered how the pattern switching emerges through the optimal pathways and how fast this occurs via the speed of pattern switching. Furthermore, both the averaged flux and the entropy production rate exhibit peaks near pattern switching boundaries, revealing dynamical and thermodynamical origins for pattern transitions, and further offering early warning signals for anticipating spatial pattern switching. Our work thus reveals physical mechanisms on spatial pattern-switching in semi-arid ecosystems and, more generally, introduces a useful approach for quantifying spatial pattern switching in nonequilibrium systems, which further offers practical applications such as early warning signals for critical transitions of spatial patterns.

As bipedal robots become more and more popular in commercial and industrial settings, the ability to control them with a high degree of reliability is critical. To that end, this paper considers how to accurately estimate which feet are currently in contact with the ground so as to avoid improper control actions that could jeopardize the stability of the robot. Additionally, modern algorithms for estimating the position and orientation of a robot's base frame rely heavily on such contact mode estimates. Dedicated contact sensors on the feet can be used to estimate this contact mode, but these sensors are prone to noise, time delays, damage/yielding from repeated impacts with the ground, and are not available on every robot. To overcome these limitations, we propose a momentum observer based method for contact mode estimation that does not rely on such contact sensors. Often, momentum observers assume that the robot's base frame can be treated as an inertial frame. However, since many humanoids' legs represent a significant portion of the overall mass, the proposed method instead utilizes multiple simultaneous dynamic models. Each of these models assumes a different contact condition. A given contact assumption is then used to constrain the full dynamics in order to avoid assuming that either the body is an inertial frame or that a fully accurate estimate of body velocity is known. The (dis)agreement between each model's estimates and measurements is used to determine which contact mode is most likely using a Markov-style fusion method. The proposed method produces contact detection accuracy of up to 98.44% with a low noise simulation and 77.12% when utilizing data collect on the Sarcos Guardian XO robot (a hybrid humanoid/exoskeleton).

Shaping the structure of light with flat optical devices has driven significant advancements in our fundamental understanding of light and light-matter interactions, and enabled a broad range of applications, from image processing and microscopy to optical communication, quantum information processing, and the manipulation of microparticles. Yet, pushing the boundaries of structured light beyond the linear optical regime remains an open challenge. Nonlinear optical interactions, such as wave mixing in nonlinear flat optics, offer a powerful platform to unlock new degrees of freedom and functionalities for generating and detecting structured light. In this study, we experimentally demonstrate the non-trivial structuring of third-harmonic light enabled by the addition of total angular momentum projection in a nonlinear, isotropic flat optics element -- a single thin film of amorphous silicon. We identify the total angular momentum projection and helicity as the most critical properties for analyzing the experimental results. The theoretical model we propose, supported by numerical simulations, offers quantitative predictions for light structuring through nonlinear wave mixing under various pumping conditions, including vectorial and non-paraxial pump light. Notably, we reveal that the shape of third-harmonic light is highly sensitive to the polarization state of the pump. Our findings demonstrate that harnessing the addition of total angular momentum projection in nonlinear wave mixing can be a powerful strategy for generating and detecting precisely controlled structured light.

Integrating watermarking into the generation process of latent diffusion models (LDMs) simplifies detection and attribution of generated content. Semantic watermarks, such as Tree-Rings and Gaussian Shading, represent a novel class of watermarking techniques that are easy to implement and highly robust against various perturbations. However, our work demonstrates a fundamental security vulnerability of semantic watermarks. We show that attackers can leverage unrelated models, even with different latent spaces and architectures (UNet vs DiT), to perform powerful and realistic forgery attacks. Specifically, we design two watermark forgery attacks. The first imprints a targeted watermark into real images by manipulating the latent representation of an arbitrary image in an unrelated LDM to get closer to the latent representation of a watermarked image. We also show that this technique can be used for watermark removal. The second attack generates new images with the target watermark by inverting a watermarked image and re-generating it with an arbitrary prompt. Both attacks just need a single reference image with the target watermark. Overall, our findings question the applicability of semantic watermarks by revealing that attackers can easily forge or remove these watermarks under realistic conditions., Comment: 23 pages, 21 figures, 6 tables

Motivated by the recent observations of various exotic quantum states in the equilateral triangular-lattice phosphates Na$_2$BaCo(PO$_4$)$_2$ with $J\rm_{eff}$ = 1/2 and Na$_2$BaNi(PO$_4$)$_2$ with $S$ = 1, the magnetic properties of spin-5/2 antiferromagnet Na$_2$BaMn(PO$_4$)$_2$, their classical counterpart, are comprehensively investigated experimentally. DC magnetization and specific heat measurements on polycrystalline samples indicate two successive magnetic transitions at $T\rm_{N1}$ $\approx$ 1.13 K and $T\rm_{N2}$ $\approx$ 1.28 K, respectively. Zero-field neutron powder diffraction measurement at 67 mK reveals a Y-like spin configuration as its ground-state magnetic structure, with both the $ab$-plane and $c$-axis components of the Mn$^{2+}$ moments long-range ordered. The incommensurate magnetic propagation vector $k$ shows a dramatic change for the intermediate phase between $T\rm_{N1}$ and $T\rm_{N2}$, in which the spin state is speculated to change into a collinear structure with only the $c$-axis moments ordered, as stabilized by thermal fluctuations. The successive magnetic transitions observed in Na$_2$BaMn(PO$_4$)$_2$ are in line with the expectation for a triangle-lattice antiferromagnet with an easy-axis magnetic anisotropy., Comment: 8 Pages, 6 figures

We saturate a high-school-level hacking benchmark with plain LLM agent design. Concretely, we obtain 95% performance on InterCode-CTF, a popular offensive security benchmark, using prompting, tool use, and multiple attempts. This beats prior work by Phuong et al. 2024 (29%) and Abramovich et al. 2024 (72%). Our results suggest that current LLMs have surpassed the high school level in offensive cybersecurity. Their hacking capabilities remain underelicited: our ReAct&Plan prompting strategy solves many challenges in 1-2 turns without complex engineering or advanced harnessing.