Search

The late time behaviour of OTOCs involving generic non conserved local operators show exponential decay in chaotic many body systems. However, it has been recently observed that for certain holographic theories, the OTOC involving the $U(1)$ conserved operator for a gauge field instead varies diffusively at late times. The present work generalizes this observation to conserved operators corresponding to higher-form symmetries that belong to a wider class of symmetries known as generalized symmetries. We started by computing the late time behaviour of OTOCs involving $U(1)$ current operators in five dimensional AdS-Schwarzschild black hole geometry for the 2-form antisymmetric $B$-fields. The bulk solution for the $B$-field exhibits logarithmic divergences near the asymptotic AdS boundary which can be regularized by introducing a double trace deformation in the boundary CFT. Finally, we consider the more general case with antisymmetric $p$-form fields in arbitrary dimensions. In the scattering approach, the boundary OTOC can be written as an inner product between asymptotic `in' and `out' states which in our case is equivalent to computing the inner product between two bulk fields with and without a shockwave background. We observe that the late time OTOC have power law tails which seems to be a universal feature of the higher-form fields with $U(1)$ charge conservation., Comment: 29 pages, 2 figures, 2 appendices

Medical imaging is spearheading the AI transformation of healthcare. Performance reporting is key to determine which methods should be translated into clinical practice. Frequently, broad conclusions are simply derived from mean performance values. In this paper, we argue that this common practice is often a misleading simplification as it ignores performance variability. Our contribution is threefold. (1) Analyzing all MICCAI segmentation papers (n = 221) published in 2023, we first observe that more than 50% of papers do not assess performance variability at all. Moreover, only one (0.5%) paper reported confidence intervals (CIs) for model performance. (2) To address the reporting bottleneck, we show that the unreported standard deviation (SD) in segmentation papers can be approximated by a second-order polynomial function of the mean Dice similarity coefficient (DSC). Based on external validation data from 56 previous MICCAI challenges, we demonstrate that this approximation can accurately reconstruct the CI of a method using information provided in publications. (3) Finally, we reconstructed 95% CIs around the mean DSC of MICCAI 2023 segmentation papers. The median CI width was 0.03 which is three times larger than the median performance gap between the first and second ranked method. For more than 60% of papers, the mean performance of the second-ranked method was within the CI of the first-ranked method. We conclude that current publications typically do not provide sufficient evidence to support which models could potentially be translated into clinical practice., Comment: Paper accepted at MICCAI 2024 conference

Efficiently evaluating the performance of text-to-image models is difficult as it inherently requires subjective judgment and human preference, making it hard to compare different models and quantify the state of the art. Leveraging Rapidata's technology, we present an efficient annotation framework that sources human feedback from a diverse, global pool of annotators. Our study collected over 2 million annotations across 4,512 images, evaluating four prominent models (DALL-E 3, Flux.1, MidJourney, and Stable Diffusion) on style preference, coherence, and text-to-image alignment. We demonstrate that our approach makes it feasible to comprehensively rank image generation models based on a vast pool of annotators and show that the diverse annotator demographics reflect the world population, significantly decreasing the risk of biases.

The paper explores "uniqueness bias," a behavioral bias defined as the tendency of planners and managers to see their decisions as singular. For the first time, uniqueness bias is correlated with forecasting accuracy and performance in real-world project investment decisions. We problematize the conventional framing of projects as unique and hypothesize that it leads to poor project performance. We test the thesis for a sample of 219 projects and find that perceived uniqueness is indeed highly statistically significantly associated with underperformance. Finally, we identify how decision makers can mitigate uniqueness bias in their projects through what Daniel Kahneman aptly called "decision hygiene," specifically reference class forecasting, premortems, similarity-based forecasting, and noise audits., Comment: 30 pages

We study the problem of fairly allocating indivisible goods among agents which are equipped with {\em leveled} valuation functions. Such preferences, that have been studied before in economics and fair division literature, capture a simple and intuitive economic behavior; larger bundles are always preferred to smaller ones. We provide a fine-grained analysis for various subclasses of leveled valuations focusing on two extensively studied notions of fairness, (approximate) MMS and EFX. In particular, we present a general positive result, showing the existence of $2/3$-MMS allocations under valuations that are both leveled and submodular. We also show how some of our ideas can be used beyond the class of leveled valuations; for the case of two submodular (not necessarily leveled) agents we show that there always exists a $2/3$-MMS allocation, complementing a recent impossibility result. Then, we switch to the case of subadditive and fractionally subadditive leveled agents, where we are able to show tight (lower and upper) bounds of $1/2$ on the approximation factor of MMS. Moreover, we show the existence of exact EFX allocations under general leveled valuations via a simple protocol that in addition satisfies several natural economic properties. Finally, we take a mechanism design approach and we propose protocols that are both truthful and approximately fair under leveled valuations.

Our work revisits the design of mechanisms via the learning-augmented framework. In this model, the algorithm is enhanced with imperfect (machine-learned) information concerning the input, usually referred to as prediction. The goal is to design algorithms whose performance degrades gently as a function of the prediction error and, in particular, perform well if the prediction is accurate, but also provide a worst-case guarantee under any possible error. This framework has been successfully applied recently to various mechanism design settings, where in most cases the mechanism is provided with a prediction about the types of the players. We adopt a perspective in which the mechanism is provided with an output recommendation. We make no assumptions about the quality of the suggested outcome, and the goal is to use the recommendation to design mechanisms with low approximation guarantees whenever the recommended outcome is reasonable, but at the same time to provide worst-case guarantees whenever the recommendation significantly deviates from the optimal one. We propose a generic, universal measure, which we call quality of recommendation, to evaluate mechanisms across various information settings. We demonstrate how this new metric can provide refined analysis in existing results. This model introduces new challenges, as the mechanism receives limited information comparing to settings that use predictions about the types of the agents. We study, through this lens, several well-studied mechanism design paradigms, devising new mechanisms, but also providing refined analysis for existing ones, using as a metric the quality of recommendation. We complement our positive results, by exploring the limitations of known classes of strategyproof mechanisms that can be devised using output recommendation.

Many accretion-powered pulsars rotate in magnetocentrifugal disequilibrium, spinning up or down secularly over multi-year intervals. The magnetic dipole moment $\mu$ of such systems cannot be inferred uniquely from the time-averaged aperiodic X-ray flux $\langle L(t) \rangle$ and pulse period $\langle P(t) \rangle$, because the radiative efficiency of the accretion is unknown and degenerate with the mass accretion rate. Here we circumvent the degeneracy by tracking the fluctuations in the unaveraged time series $L(t)$ and $P(t)$ using an unscented Kalman filter, whereupon $\mu$ can be estimated uniquely, up to the uncertainties in the mass, radius and distance of the star. The analysis is performed on Rossi X-ray Timing Explorer observations for $24$ X-ray transients in the Small Magellanic Cloud, which have been monitored regularly for $\sim 16$ years. As well as independent estimates of $\mu$, the analysis yields time-resolved histories of the mass accretion rate and the Maxwell stress at the disk-magnetosphere boundary for each star, and hence auto- and cross-correlations involving the latter two state variables. The inferred fluctuation statistics convey important information about the complex accretion physics at the disk-magnetosphere boundary., Comment: 42 pages, 10 figures, accepted for publication in The Astrophysical Journal

This study delves into [W60] B90, one of the most luminous and extreme Red Supergiants (RSGs) in the Large Magellanic Cloud (LMC), aiming to search for evidence of episodic mass loss. Our discovery of a bar-like nebular structure at 1 pc, reminiscent of the bar around Betelgeuse, raised the question of whether [W60] B90 also has a bow shock. We collected and analyzed proper motion data from Gaia, as well as new multi-epoch spectroscopic and imaging data, and archival time-series photometry in the optical and mid-infrared. We found [W60] B90 to be a walkaway star, with a supersonic peculiar velocity in the direction of the bar. We detected shocked emission between the bar and the star, based on the [S II]/H$\alpha$ > 0.4 criterion, providing strong evidence for a bow shock. The 30-year optical light curve revealed semi-regular variability, showing three similar dimming events with $\Delta V \sim 1$ mag, a recurrence of $\sim$12 yr, and a rise time of 400 d. We found the mid-IR light curve to vary by 0.51 mag and 0.37 mag in the WISE1 and WISE2 bands, respectively, and by 0.42 mag and 0.25 mag during the last dimming event. During this event, optical spectroscopy revealed spectral variability (M3I to M4I), a correlation between the $T_{\rm eff}$ and the brightness, increased extinction, and, after the minimum, spectral features incompatible with the models. We also found a difference of >300 K between the $T_{\rm eff}$ measured from the TiO bands in the optical and the atomic lines from our $J$-band spectroscopy. We inferred that [W60] B90 is a more massive analog of Betelgeuse in the LMC and the first extragalactic single RSG with a suspected bow shock. Its high luminosity $\log(L/L_{\odot})=5.32$ dex, mass-loss rate, and mid-IR variability compared to other RSGs in the LMC, indicate that it is in an unstable evolutionary state undergoing episodes of mass loss., Comment: 28 pages, 14 figures, submitted to A&A

Evolutionary Algorithms and Generative Deep Learning have been two of the most powerful tools for sound generation tasks. However, they have limitations: Evolutionary Algorithms require complicated designs, posing challenges in control and achieving realistic sound generation. Generative Deep Learning models often copy from the dataset and lack creativity. In this paper, we propose LVNS-RAVE, a method to combine Evolutionary Algorithms and Generative Deep Learning to produce realistic and novel sounds. We use the RAVE model as the sound generator and the VGGish model as a novelty evaluator in the Latent Vector Novelty Search (LVNS) algorithm. The reported experiments show that the method can successfully generate diversified, novel audio samples under different mutation setups using different pre-trained RAVE models. The characteristics of the generation process can be easily controlled with the mutation parameters. The proposed algorithm can be a creative tool for sound artists and musicians., Comment: Accepted to GECCO 24 Companion

Experimental proposals for testing quantum gravity-induced entanglement of masses (QGEM) typically involve two interacting masses which are each in a spatial superposition state. Here, we propose a QGEM experiment with two particles which are each in a superposition of rotational states, this amounts to a superposition of mass through mass-energy equivalence. Our proposal relies on the fact that rotational energy gravitates. This approach would test a feature unique to gravity since it amounts to sourcing a spacetime in superposition due to a superposition of 'charge'. We propose and analyse a concrete experimental protocol and discuss challenges., Comment: 5 pages, 3 figures, 3 pages supplemental material

Mass loss during the red supergiant (RSG) phase plays a crucial role in the evolution of an intermediate massive star, however, the underlying mechanism remains unknown. We aim to increase the sample of well-characterized RSGs at subsolar metallicity, by deriving the physical properties of 127 RSGs in nine nearby, southern galaxies presented by Bonanos et al. For each RSG, we provide spectral types and used \textsc{marcs} atmospheric models to measure stellar properties from their optical spectra, such as the effective temperature, extinction, and radial velocity. By fitting the spectral energy distribution, we obtained the stellar luminosity and radius for 97 RSGs, finding $\sim 50\%$ with log$(L/ \rm L_{\odot}) \geq 5.0$ and 6 RSGs with $R \gtrsim 1400 \,\ \rm R_{\odot}$. We also find a correlation between the stellar luminosity and mid-IR excess of 33 dusty, variable sources. Three of these dusty RSGs have luminosities exceeding the revised Humphreys-Davidson limit. We then derive a metallicity-dependent $J-K_s$ color versus temperature relation from synthetic photometry and two new empirical $J-K_s$ color versus temperature relations calibrated on literature TiO and $J$-band temperatures. To scale our derived, cool TiO temperatures to values in agreement with the evolutionary tracks, we derive two linear scaling relations calibrated on $J$-band and $i$-band temperatures. We find that the TiO temperatures are more discrepant as a function of the mass-loss rate and discuss future prospects of the TiO bands as a mass-loss probe. Finally, we speculate that 3 hot, dusty RSGs may have experienced a recent mass ejection ($12\%$ of the K-type sample) and indicate them as candidate Levesque-Massey variables., Comment: 22 pages, 18 figures, submitted to A&A

Self-projective sets are natural fractal sets which describe the action of a semigroup of matrices on projective space. In recent years there has been growing interest in studying the dimension theory of self-projective sets, as well as progress in the understanding of closely related objects such as Furstenberg measures. The aim of this survey is twofold: first to motivate the study of these objects from several different perspectives and second to make the study of these objects more accessible for readers with expertise in iterated function systems., Comment: Survey submitted to the Proceedings of Fractals and Related Fields IV conference

The magnetic dipole moment $\mu$ of an accretion-powered pulsar in magnetocentrifugal equilibrium cannot be inferred uniquely from time-averaged pulse period and aperiodic X-ray flux data, because the radiative efficiency $\eta_0$ of the accretion is unknown, as are the mass, radius, and distance of the star. The degeneracy associated with the radiative efficiency is circumvented, if fluctuations of the pulse period and aperiodic X-ray flux are tracked with a Kalman filter, whereupon $\mu$ can be measured uniquely up to the uncertainties in the mass, radius, and distance. Here the Kalman filter analysis is demonstrated successfully in practice for the first time on Rossi X-ray Timing Explorer observations of the X-ray transient SXP 18.3 in the Small Magellanic Cloud, which is monitored regularly. The analysis yields $\mu = 8.0^{+1.3}_{-1.2} \, \times \, 10^{30} \, {\rm G \, cm^3}$ and $\eta_0 = 0.04^{+0.02}_{-0.01}$, compared to $\mu = 5.0^{+1.0}_{-1.0} \times 10^{30} \, {\rm G \, cm^3}$ as inferred traditionally from time-averaged data assuming $\eta_0=1$. The analysis also yields time-resolved estimates of two hidden state variables, the mass accretion rate and the Maxwell stress at the disk-magnetosphere boundary. The success of the demonstration confirms that the Kalman filter analysis can be applied in the future to study the magnetic moments and disk-magnetosphere physics of accretion-powered pulsar populations in the Small Magellanic Cloud and elsewhere., Comment: 34 pages, 5 figures, accepted for publication in The Astrophysical Journal

We report the observations of two self-lensing pulses from KIC 12254688 in Transiting Exoplanet Survey Satellite (TESS) light curves. This system, containing a F2V star and white-dwarf companion, was amongst the first self-lensing binary systems discovered by the Kepler Space Telescope over the past decade. Each observed pulse occurs when the white dwarf transits in front of its companion star, gravitationally lensing the star's surface, thus making it appear brighter to a distant observer. These two pulses are the very first self-lensing events discovered in TESS observations. We describe the methods by which the data were acquired and detrended, as well as the best-fit binary parameters deduced from our self-lensing+radial velocity model. We highlight the difficulties of finding new self-lensing systems with TESS, and we discuss the types of self-lensing systems that TESS may be more likely to discover in the future., Comment: Accepted for publication in ApJ Letters, 12 Pages, 6 Figures, and 1 Table

Developing stateful cloud applications, such as low-latency workflows and microservices with strict consistency requirements, remains arduous for programmers. The Stateful Functions-as-a-Service (SFaaS) paradigm aims to serve these use cases. However, existing approaches either provide serializable transactional guarantees at the level of individual functions, or separate application logic from the state and use inefficient transactional protocols. These design choices increase the execution latency, limiting the adoption of SFaaS systems. In this paper, we present Styx, a novel SFaaS runtime that executes serializable transactions across functions with exactly-once guarantees. Styx extends a deterministic transactional protocol to support an arbitrary call graph of stateful functions. It introduces a transaction-execution acknowledgment scheme that allows tracking a transactional workflow's SFaaS calls, guaranteeing atomicity and exactly-once processing. Finally, Styx features a function-execution caching mechanism and early transactional commit replies for optimized performance. Experiments with the YCSB-T, TPC-C, and Deathstar benchmarks show that Styx outperforms state-of-the-art approaches by achieving at least one order of magnitude higher throughput while exhibiting near-linear scalability and low latency.

The ASSESS project aims to determine the role of episodic mass-loss in the evolution of massive stars. As a first step, we construct a catalog of spectroscopically identified dusty, evolved massive stars in ten southern galaxies for which Spitzer point-source catalogs are available. We conducted multi-object spectroscopy of dusty massive star candidates in these galaxies (spanning Z = 0.06-1.6 Zo) using the VLT. We obtained 763 spectra in WLM, NGC 55, NGC 247, NGC 253, NGC 300, NGC 1313, NGC 3109, Sextans A, M83 and NGC 7793. The targets were selected using their Spitzer photometry, by prioritizing targets with a strong infrared excess. We determined a spectral classification for each target. Additionally, we used archival images from the HST to provide a visual classification for 80 targets, as a star, cluster, or galaxy. We provide a catalog of 541 spectroscopically classified sources including 185 massive stars, of which 154 are newly classified massive stars. The catalog contains 129 red supergiants, 27 blue supergiants, 10 yellow supergiants, four luminous blue variable candidates, seven supergiant B[e] stars and eight emission line objects. Evidence for circumstellar dust is found in 24% of these massive stars, based on their infrared colors. We report a success rate of 28% for identifying massive stars among our observed spectra, while the average success rate of our priority system in selecting evolved massive stars was 36%. Additionally, the catalog contains 21 background galaxies (including AGN and quasars), 10 carbon stars and 99 HII regions. We measured the line ratios [NII]/Ha and [SII]/Ha for 76 HII regions and 36 other spectra with nebular emission-lines, thereby identifying eight sources with shocked emission. We present the largest catalog of evolved massive stars and in particular of red supergiants in nearby galaxies at low Z beyond the Local Group., Comment: 47 pages, 7 figures, 9 tables, 5 appendices. Accepted for publication in A&A

The task of predicting 3D eye gaze from eye images can be performed either by (a) end-to-end learning for image-to-gaze mapping or by (b) fitting a 3D eye model onto images. The former case requires 3D gaze labels, while the latter requires eye semantics or landmarks to facilitate the model fitting. Although obtaining eye semantics and landmarks is relatively easy, fitting an accurate 3D eye model on them remains to be very challenging due to its ill-posed nature in general. On the other hand, obtaining large-scale 3D gaze data is cumbersome due to the required hardware setups and computational demands. In this work, we propose to predict 3D eye gaze from weak supervision of eye semantic segmentation masks and direct supervision of a few 3D gaze vectors. The proposed method combines the best of both worlds by leveraging large amounts of weak annotations--which are easy to obtain, and only a few 3D gaze vectors--which alleviate the difficulty of fitting 3D eye models on the semantic segmentation of eye images. Thus, the eye gaze vectors, used in the model fitting, are directly supervised using the few-shot gaze labels. Additionally, we propose a transformer-based network architecture, that serves as a solid baseline for our improvements. Our experiments in diverse settings illustrate the significant benefits of the proposed method, achieving about 5 degrees lower angular gaze error over the baseline, when only 0.05% 3D annotations of the training images are used. The source code is available at https://github.com/dimitris-christodoulou57/Model-aware_3D_Eye_Gaze., Comment: Accepted to ISMAR2023 as a poster paper

The Medipix4 chip is the latest member in the Medipix/Timepix family of hybrid pixel detector chips aimed at high-rate spectroscopic X-ray imaging using high-Z materials. It can be tiled on all 4 sides making it ideal for constructing large-area detectors with minimal dead area. The chip is designed to read out a sensor of 320 x 320 pixels with dimensions of 75 {\mu}m x 75 {\mu}m or 160 x 160 pixels with dimensions of 150 {\mu}m x 150 {\mu}m. The readout architecture features energy binning of the single photons, which includes charge sharing correction for hits with energy spread over adjacent pixels. This paper presents the specifications, architecture, and circuit implementation of the chip, along with the first electrical measurements.

Protocols for observing gravity induced entanglement typically comprise the interaction of two particles prepared either in a superposition of two discrete paths, or in a continuously delocalized (harmonic oscillator) state of motion. An important open question has been whether these two different approaches allow to draw the same conclusions on the quantum nature of gravity. To answer this question, we analyse using the path-integral approach a setup that contains both features: a superposition of two highly delocalized center of mass states. We conclude that the two usual protocols are of similar epistemological relevance. In both cases the appearance of entanglement, within linearised quantum gravity, is due to gravity being in a highly non-classical state: a superposition of distinct geometries., Comment: 5 + 2 pages, + Mathematica Notebooks

The use of colloidal quantum dots (CQDs) as a gain medium in infrared laser devices has been underpinned by the need for high pumping intensities, very short gain lifetimes and low gain coefficients.

In spite or the large astronomical evidence for its existence, the nature of dark matter remains enigmatic. Particles that interact only, or almost only, gravitationally, in particular with masses around the Planck mass -- the fundamental scale in quantum gravity, are intriguing candidates. Here we show that there is a theoretical possibility to directly detect such particles using highly sensitive gravity-mediated quantum phase shifts. In particular, we consider a protocol utilizing Josephson junctions., Comment: A refined protocol using Josephson junctions has been added

We report an updated measurement of the $\nu_{\mu}$-induced, and the first measurement of the $\bar{\nu}_{\mu}$-induced coherent charged pion production cross section on $^{12}C$ nuclei in the T2K experiment. This is measured in a restricted region of the final-state phase space for which $p_{\mu,\pi} > 0.2$ GeV, $\cos(\theta_{\mu}) > 0.8$ and $\cos(\theta_{\pi}) > 0.6$, and at a mean (anti)neutrino energy of 0.85 GeV using the T2K near detector. The measured $\nu_{\mu}$ CC coherent pion production flux-averaged cross section on $^{12}C$ is $(2.98 \pm 0.37 (stat.) \pm 0.31 (syst.) \substack{ +0.49 \\ -0.00 } \mathrm{ (Q^2\,model)}) \times 10^{-40}~\mathrm{cm}^{2}$. The new measurement of the $\bar{\nu}_{\mu}$-induced cross section on $^{12}{C}$ is $(3.05 \pm 0.71 (stat.) \pm 0.39 (syst.) \substack{ +0.74 \\ -0.00 } \mathrm{(Q^2\,model)}) \times 10^{-40}~\mathrm{cm}^{2}$. The results are compatible with both the NEUT 5.4.0 Berger-Sehgal (2009) and GENIE 2.8.0 Rein-Sehgal (2007) model predictions.

Episodic mass loss is not understood theoretically, neither accounted for in state-of-the-art models of stellar evolution, which has far-reaching consequences for many areas of astronomy. We introduce the ERC-funded ASSESS project (2018-2024), which aims to determine whether episodic mass loss is a dominant process in the evolution of the most massive stars, by conducting the first extensive, multi-wavelength survey of evolved massive stars in the nearby Universe. It hinges on the fact that mass-losing stars form dust and are bright in the mid-infrared. We aim to derive physical parameters of $\sim$1000 dusty, evolved massive stars in $\sim$25 nearby galaxies and estimate the amount of ejected mass, which will constrain evolutionary models, and quantify the duration and frequency of episodic mass loss as a function of metallicity. The approach involves applying machine-learning algorithms to select dusty, luminous targets from existing multi-band photometry of nearby galaxies. We present the first results of the project, including the machine-learning methodology for target selection and results from our spectroscopic observations so far. The emerging trend for the ubiquity of episodic mass loss, if confirmed, will be key to understanding the explosive early Universe and will have profound consequences for low-metallicity stars, reionization, and the chemical evolution of galaxies., Comment: To appear in the proceedings of IAU Symposium 361: Massive stars near and far; 6 pages, 1 figure