Search

This paper explores the attributes necessary to determine the creative merit of intermediate artifacts produced during a computational co-creative process (CCC) in which a human and an artificial intelligence system collaborate in the generative phase of a creative project. In an active listening experiment, subjects with diverse musical training (N=43) judged unfinished pieces composed by the New Electronic Assistant (NEA). The results revealed that a two-attribute definition based on the value and novelty of an artifact (e.g., Corazza's effectiveness and novelty) suffices to assess unfinished work leading to innovative products, instead of Boden's classic three-attribute definition of creativity (value, novelty, and surprise). These findings reduce the creativity metrics needed in CCC processes and simplify the evaluation of the numerous unfinished artifacts generated by computational creative assistants., Comment: Conference proceedings paper

Let G be a locally compact group and let $\phi$ be a positive definite function on G with $\phi(e)=1$. This function defines a multiplication operator $M_\phi$ on the Fourier algebra $A(G)$ of $G$. The aim of this paper is to classify the ergodic properties of the operators $M_\phi$, focusing on several key factors, including the subgroup $H_\phi=\{x\in G\colon \phi(x)=1\}$, the spectrum of $M_\phi$, or how ``spread-out'' a power of $M_\phi$ can be. We show that the multiplication operator $M_\phi$ is uniformly mean ergodic if and only if $H_\phi$ is open and 1 is not an accumulation point of the spectrum of $M_\phi$. Equivalently, this happens when some power of $\phi$ is not far, in the multiplier norm, from a function supported on finitely many cosets of $H_\phi$. Additionally, we show that the powers of $M_\phi$ converge in norm if, and only if, the operator is uniformly mean ergodic and $H_\phi =\{x\in G\colon |\phi(x)|=1\}$., Comment: 21 pages

We employ a flexible parametric model to estimate global income, health, and education distributions from 1980 to 2015. Using these marginal distributions within a copula-based framework, we construct a global joint distribution of well-being. This approach allows us to specifically analyze the impact of dependency structures on global well-being inequality. While inequality decreased in each individual dimension, our findings suggest that multidimensional inequality does not necessarily follow this trend. Its evolution is influenced by the interdependence among dimensions and the chosen inequality aversion parameter.

We provide a new gravity field of Comet 67P-C/G up to degree 4. We detect mass heterogeneity in the comet nucleus. The loss of mass is restimated at 0.28\% of the comet's total mass (3 times larger than previous estimate). Comparison of the gravity field between pre- and post-perihelion allowed us to measure a shift in the comet's center of gravity of 35 m northward, attributed to ice sublimation process.

The behaviour of the generalized Hilbert operator associated with a positive finite Borel measure $\mu$ on $[0,1)$ is investigated when it acts on weighted Banach spaces of holomorphic functions on the unit disc defined by sup-norms and on Korenblum type growth Banach spaces. It is studied when the operator is well defined, bounded and compact. To this aim, we study when it can be represented as an integral operator. We observe important differences with the behaviour of the Ces\`aro-type operator acting on these spaces, getting that boundedness and compactness are equivalent concepts for some standard weights. For the space of bounded holomorphic functions on the disc and for the Wiener algebra, we get also this equivalence, which is characterized in turn by the summability of the moments of the measure $\mu.$ In the latter case, it is also equivalent to nuclearity. Nuclearity of the generalized Hilbert operator acting on related spaces, such as the classical Hardy space, is also analyzed.

Stimuli-responsive macromolecules display large conformational changes during their dynamics, sometimes switching states, and are an integral property for the development of soft functional materials. Here, we introduce a mean-field dynamical density functional theory (DDFT) for a model of responsive colloids (RCs) to study the nonequilibrium dynamics of a colloidal dispersion in time-dependent external fields, with a focus on the coupling of translational and conformational dynamics during their relaxation. Specifically, we consider soft Gaussian particles with a bimodal size distribution between two confining walls with time-dependent (switching-on and off) external gravitational and osmotic fields. We find a rich relaxation behavior of the systems in excellent agreement with particle-based Brownian dynamics (BD) computer simulations. In particular, we find time-asymmetric relaxations of integrated observables (wall pressures, mean size, liquid center-of-mass) for activation/deactivation of external potentials, respectively, which are tuneable by the ratio of translational and conformational diffusion time scales. Our work paves thus the way for studying the nonequilibrium relaxation dynamics of complex soft matter with multiple degrees of freedom and hierarchical relaxations., Comment: 14 pages, 8 figures

The present work aims to determine the source regions of carbonaceous chondrites (CM, CI, CO, CV, CK, CR, CH, CB, or C-ungrouped). We studied 38 individual asteroid families, including young and old ones, and determined their contributions to the NEO populations at metre and kilometre sizes using collisional and orbital models. Our models are in agreement with spectroscopic observations of NEOs, cosmic-ray exposure ages of meteorites, statistics of bolides, infrared emission from dust bands, composition of interplanetary dust particles (IDPs), or abundance of extraterrestrial helium-3. We identified the Veritas, Polana and Eos families as the primary sources of CM/CR, CI and CO/CV/CK chondrites, respectively. Substantial contributions are also expected from CM-like K\"onig and CI-like Clarissa, Misa and Hoffmeister families. The source regions of kilometre-sized bodies are generally different. The Adeona family is by far the main source of CM-like NEOs, whereas the Polana (low-i) and Euphrosyne (high-i) families are at the origin of most CI-like NEOs. The Polana family is the likely source of both Ryugu and Bennu. We were able to link spectroscopically and dynamically several NEOs to the Baptistina family. Finally, it appears that the pre-atmospheric flux of carbonaceous chondrites at metre sizes is about the same as that of ordinary chondrites. Given the difference in fall statistics between the two groups (80\% versus 4.4\%), this implies either substantial atmospheric fragmentation of carbonaceous bodies at the level of ${\sim}0.5\,{\rm MPa}$, or destruction by thermal cracking and water desorption. The source regions of most meteorites and kilometre-sized NEOs have now been determined; including some minor classes like enstatite chondrites and achondrites (Nysa, Hungaria), acapulcoites/lodranites (Iannini)., Comment: 47 pages, 31 figures, accepted in A&A

We present our work on developing and training scalable, trustworthy, and energy-efficient predictive graph foundation models (GFMs) using HydraGNN, a multi-headed graph convolutional neural network architecture. HydraGNN expands the boundaries of graph neural network (GNN) computations in both training scale and data diversity. It abstracts over message passing algorithms, allowing both reproduction of and comparison across algorithmic innovations that define nearest-neighbor convolution in GNNs. This work discusses a series of optimizations that have allowed scaling up the GFMs training to tens of thousands of GPUs on datasets consisting of hundreds of millions of graphs. Our GFMs use multi-task learning (MTL) to simultaneously learn graph-level and node-level properties of atomistic structures, such as energy and atomic forces. Using over 154 million atomistic structures for training, we illustrate the performance of our approach along with the lessons learned on two state-of-the-art United States Department of Energy (US-DOE) supercomputers, namely the Perlmutter petascale system at the National Energy Research Scientific Computing Center and the Frontier exascale system at Oak Ridge Leadership Computing Facility. The HydraGNN architecture enables the GFM to achieve near-linear strong scaling performance using more than 2,000 GPUs on Perlmutter and 16,000 GPUs on Frontier., Comment: 51 pages, 32 figures

We characterize those pairs $(\psi,\varphi)$ of smooth mappings $\psi:\mathbb{R}^d\rightarrow\mathbb{C},\varphi:\mathbb{R}^d\rightarrow\mathbb{R}^d$ for which the corresponding weighted composition operator $C_{\psi,\varphi}f=\psi\cdot(f\circ\varphi)$ acts continuously on $\mathscr{S}(\mathbb{R}^d)$. Additionally, we give several easy-to-check necessary and sufficient conditions of this property for interesting special cases. Moreover, we characterize power boundedness and topologizablity of $C_{\psi,\varphi}$ on $\mathscr{S}(\mathbb{R}^d)$ in terms of $\psi,\varphi$. Among other things, as an application of our results we show that for a univariate polynomial $\varphi$ with $\text{deg}(\varphi)\geq 2$, power boundedness of $C_{\psi,\varphi}$ on $\mathscr{S}(\mathbb{R})$ for every $\psi\in\mathscr{O}_M(\mathbb{R})$ only depends on $\varphi$ and that in this case power boundedness of $C_{\psi,\varphi}$ is equivalent to $(C_{\psi,\varphi}^n)_{n\in\mathbb{N}}$ converging to $0$ in $\mathcal{L}_b(\mathscr{S}(\mathbb{R}))$ as well as to the uniform mean ergodicity of $C_{\psi,\varphi}$. Additionally, we give an example of a power bounded and uniformly mean ergodic weighted composition operator $C_{\psi,\varphi}$ on $\mathscr{S}(\mathbb{R})$ for which neither the multiplication operator $f\mapsto \psi f$ nor the composition operator $f\mapsto f\circ\varphi$ acts on $\mathscr{S}(\mathbb{R})$. Our results complement and considerably extend various results of Fern\'andez, Galbis, and the second named author., Comment: 22 pages; comments welcome; minor editorial changes; accepted for publication in Journal of Functional Analysis

Single-detector 3D dynamic light scattering (3D-DLS) emerges as a reliable technique to determine the drift velocity of out-of-equilibrium colloidal particles. In particular, our investigation reveals the appearance of oscillations of a well-defined frequency in the autocorrelation function of the scattered intensity when particles are immersed in a medium exposed to thermally induced convection. These oscillations arise as a consequence of the directed motion of particles due to the convective motion of the fluid. The experimental results obtained for different colloidal systems are corroborated by a theoretical model and thoroughly validated with fluid dynamics and Brownian dynamics simulations. The excellent agreement between experimental, theoretical and simulation data allows us to provide a solid and comprehensive explanation of the observed physical phenomena. This study via advanced dynamic light scattering (DLS) technique offers insights into the field of non-equilibrium particle dynamics, applicable not only to colloidal suspension affected by steady-state diffusion-convection but also to other non-equilibrium situations, such as systems driven by external fields (gravitational, electric or magnetic fields, among others)., Comment: 5 pages, 4 figures

Let $\mu$ be a positive finite Borel measure on $[0,1).$ Ces\`aro-type operators $C_{\mu}$ when acting on weighted spaces of holomorphic functions are investigated. In the case of bounded holomorphic functions on the unit disc we prove that $C_\mu$ is continuous if and only if it is compact. In the case of weighted Banach spaces of holomorphic function defined by general weights, we give sufficient and necessary conditions for the continuity and compactness. For standard weights, we characterize the continuity and compactness on classical growth Banach spaces of holomorphic functions. We also study the point spectrum and the spectrum of $C_\mu$ on the space of holomorphic functions on the disc, on the space of bounded holomorphic functions on the disc, and on the classical growth Banach spaces of holomorphic functions. All characterizations are given in terms of the sequence of moments $(\mu_n)_{n\in\N_0}$. The continuity, compactness and spectrum of $C_\mu$ acting on Fr\'echet and (LB) Korenblum type spaces are also considered.

We explore the utilization of the Apache TVM open source framework to automatically generate a family of algorithms that follow the approach taken by popular linear algebra libraries, such as GotoBLAS2, BLIS and OpenBLAS, in order to obtain high-performance blocked formulations of the general matrix multiplication (GEMM). % In addition, we fully automatize the generation process, by also leveraging the Apache TVM framework to derive a complete variety of the processor-specific micro-kernels for GEMM. This is in contrast with the convention in high performance libraries, which hand-encode a single micro-kernel per architecture using Assembly code. % In global, the combination of our TVM-generated blocked algorithms and micro-kernels for GEMM 1)~improves portability, maintainability and, globally, streamlines the software life cycle; 2)~provides high flexibility to easily tailor and optimize the solution to different data types, processor architectures, and matrix operand shapes, yielding performance on a par (or even superior for specific matrix shapes) with that of hand-tuned libraries; and 3)~features a small memory footprint., Comment: 35 pages, 22 figures. Submitted to ACM TOMS

The transport of biomolecules, drugs, or reactants encapsulated inside stimuli-responsive polymer networks in aqueous media is fundamental for many material and environmental science applications, including drug delivery, biosensing, catalysis, nanofiltration, water purification, and desalination. The transport is particularly complex in dense polymer media, such as collapsed hydrogels, where the molecules strongly interact with the polymer network and diffuse via a hopping mechanism. In this study, we employ Dynamical Density Functional Theory (DDFT) to investigate the non-equilibrium release kinetics of non-ionic subnanometer-sized molecules initially uploaded inside collapsed microgel particles. The theory is consistent with previous molecular dynamics simulations of collapsed poly($N$-isopropylacrylamide) (PNIPAM) polymer matrices, accommodating molecules of varying shapes and sizes. We found that, despite the intricate physico-chemical properties involved in the released process, the kinetics is predominantly dictated by two material parameters: the diffusion coefficient of the molecules inside the microgel ($D^*$) and the interaction free energy of the molecules with the microgel ($\Delta G$). Our results reveal two distinct limiting regimes: For large, slowly diffusing molecules weakly attracted to the polymer network, the release is primarily driven by diffusion, with a release time that scales as $\tau_{1/2} \sim 1/D^*$. Conversely, for small molecules strongly attracted to the polymer network, the release time is dominated by the interaction, scaling as $\tau_{1/2} \sim \exp(-\Delta G/k_{\textrm{B}} T)$. Our DDFT calculations are directly compared with an analytical equation for the half-release time, demonstrating excellent quantitative agreement. This equation represents a valuable tool for predicting release kinetics of non-ionic molecules from collapsed microgels., Comment: 15 pages, 6 figures

Inference for impulse responses estimated with local projections presents interesting challenges and opportunities. Analysts typically want to assess the precision of individual estimates, explore the dynamic evolution of the response over particular regions, and generally determine whether the impulse generates a response that is any different from the null of no effect. Each of these goals requires a different approach to inference. In this article, we provide an overview of results that have appeared in the literature in the past 20 years along with some new procedures that we introduce here.

Let $G$ be a locally compact group and $\mu$ be a probability measure on $G$. We consider the convolution operator $\lambda_1(\mu)\colon L_1(G)\to L_1(G)$ given by $\lambda_1(\mu)f=\mu \ast f$ and its restriction $\lambda_1^0(\mu)$ to the augmentation ideal $L_1^0(G)$. Say that $\mu$ is uniformly ergodic if the Ces\`aro means of the operator $\lambda_1^0(\mu)$ converge uniformly to 0, that is, if $\lambda_1^0(\mu)$ is a uniformly mean ergodic operator with limit 0 and that $\mu$ is uniformly completely mixing if the powers of the operator $\lambda_1^0(\mu)$ converge uniformly to 0. We completely characterize the uniform mean ergodicity of the operator $\lambda_1(\mu)$ and the uniform convergence of its powers and see that there is no difference between $\lambda_1(\mu)$ and $\lambda_1^0(\mu) $ in this regard. We prove in particular that $\mu$ is uniformly ergodic if and only if $G$ is compact, $\mu$ is adapted (its support is not contained in a proper closed subgroup of $G$) and 1 is an isolated point of the spectrum of $\mu$. The last of these three conditions is actually equivalent to $\mu$ being spread-out (some convolution power of $\mu$ is not singular). The measure $\mu$ is uniformly completely mixing if and only if $G$ is compact, $\mu$ is spread-out and the only unimodular value of the spectrum of $\mu$ is 1., Comment: Final Version. To appear in Publicacions Matem\`atiques

In this note we study the spectrum and the Waelbroeck spectrum of the derivative operator composed with isomorphic multiplication oper

Aims. Understanding the activity is vital for deciphering the structure, formation, and evolution of comets. We investigate models of cometary activity by comparing them to the dynamics of 67P/Churyumov-Gerasimenko. Methods. We matched simple thermal models of water activity to the combined Rosetta datasets by fitting to the total outgassing rate and four components of the outgassing induced non-gravitational force and torque, with a final manual adjustment of the model parameters to additionally match the other two torque components. We parametrised the thermal model in terms of a distribution of relative activity over the surface of the comet, and attempted to link this to different terrain types. We also tested a more advanced thermal model based on a pebble structure. Results. We confirm a hemispherical dichotomy and non-linear water outgassing response to insolation. The southern hemisphere of the comet and consolidated terrain show enhanced activity relative to the northern hemisphere and dust-covered, unconsolidated terrain types, especially at perihelion. We further find that the non-gravitational torque is especially sensitive to the activity distribution, and to fit the pole-axis orientation in particular, activity must be concentrated (in excess of the already high activity in the southern hemisphere and consolidated terrain) around the south pole and on the body and neck of the comet over its head. This is the case for both the simple thermal model and the pebble-based model. Overall, our results show that water activity cannot be matched by a simple model of sublimating surface ice driven by the insolation alone, regardless of the surface distribution, and that both local spatial and temporal variations are needed to fit the data., Comment: Accepted for publication in A&A. 13 pages, 22 figures including appendix

Cybele asteroids constitute an appealing reservoir of primitive material genetically linked to the outer Solar System, and the physical properties of the largest members can be readily accessed by large telescopes. We took advantage of the bright apparition of (65) Cybele in July and August 2021 to acquire high-angular-resolution images and optical light curves of the asteroid with which we aim to analyse its shape and bulk properties. 7 series of images acquired with VLT/SPHERE were combined with optical light curves to reconstruct the shape of the asteroid using the ADAM, MPCD, and SAGE algorithms. The origin of the shape was investigated by means of N-body simulations. Cybele has a volume-equivalent diameter of 263+/-3km and a bulk density of 1.55+/-0.19g.cm-3. Notably, its shape and rotation state are closely compatible with those of a Maclaurin equilibrium figure. The lack of a collisional family associated with Cybele and the higher bulk density of that body with respect to other large P-type asteroids suggest that it never experienced any large disruptive impact followed by rapid re-accumulation. This would imply that its present-day shape represents the original one. However, numerical integration of the long-term dynamical evolution of a hypothetical family shows that it is dispersed by gravitational perturbations and chaotic diffusion over Gyrs of evolution. The very close match between Cybele and an equilibrium figure opens up the possibility that D>260km small bodies from the outer Solar System all formed at equilibrium. However, we cannot rule out an old impact as the origin of the equilibrium shape. Cybele itself is found to be dynamically unstable, implying that it was recently (<1Ga) placed on its current orbit either through slow diffusion from a relatively stable orbit in the Cybele region or, less likely, from an unstable, JFC orbit in the planet-crossing region., Comment: 19 pages, 14 figures, 4 tables, accepted for publication in A&A

The Large Hadron Collider (LHC) at CERN has generated in the last decade an unprecedented volume of data for the High-Energy Physics (HEP) field. Scientific collaborations interested in analysing such data very often require computing power beyond a single machine. This issue has been tackled traditionally by running analyses in distributed environments using stateful, managed batch computing systems. While this approach has been effective so far, current estimates for future computing needs of the field present large scaling challenges. Such a managed approach may not be the only viable way to tackle them and an interesting alternative could be provided by serverless architectures, to enable an even larger scaling potential. This work describes a novel approach to running real HEP scientific applications through a distributed serverless computing engine. The engine is built upon ROOT, a well-established HEP data analysis software, and distributes its computations to a large pool of concurrent executions on Amazon Web Services Lambda Serverless Platform. Thanks to the developed tool, physicists are able to access datasets stored at CERN (also those that are under restricted access policies) and process it on remote infrastructures outside of their typical environment. The analysis of the serverless functions is monitored at runtime to gather performance metrics, both for data- and computation-intensive workloads., Comment: 10 pages, CCGRID 2022

In this paper we consider composition operators on locally convex spaces of functions defined on $\mathbb{R}$. We prove results concerning supercyclicity, power boundedness, mean ergodicity and convergence of the iterates in the strong operator topology.

We employ Reactive Dynamical Density Functional Theory (R-DDFT) and Reactive Brownian Dynamics (R-BD) simulations to investigate the dynamics of a suspension of active soft Gaussian colloids with binary interaction switching, i.e., a one-component colloidal system in which every particle stochastically switches at predefined rates between two interaction states with different mobility. Using R-DDFT we extend a theory previously developed to access the dynamics of inhomogeneous liquids to study the influence of the switching activity on the self and distinct part of the Van Hove function in bulk solution, and determine the corresponding mean squared displacement of the switching particles. Our results demonstrate that, even though the average diffusion coefficient is not affected by the switching activity, it significantly modifies the non-equilibrium dynamics and diffusion coefficients of the individual particles, leading to a crossover from short to long times, with a regime for intermediate times showing anomalous diffusion. In addition, the self-part of the van Hove function has a Gaussian form at short and long times, but becomes non-Gaussian at intermediates ones, having a crossover between short and large displacements. The corresponding self-intermediate scattering function shows the two-step relaxation patters typically observed in soft materials with heterogeneous dynamics such as glasses and gels. We also introduce a phenomenological Continuous Time Random Walk (CTRW) theory to understand the heterogeneous diffusion of this system. R-DDFT results are in excellent agreement with R-BD simulations and the analytical predictions of CTRW theory, thus confirming that R-DDFT constitutes a powerful method to investigate not only the structure and phase behavior, but also the dynamical properties of non-equilibrium active switching colloidal suspensions., Comment: 15 pages, 7 figures

As the number of installed cameras grows, so do the compute resources required to process and analyze all the images captured by these cameras. Video analytics enables new use cases, such as smart cities or autonomous driving. At the same time, it urges service providers to install additional compute resources to cope with the demand while the strict latency requirements push compute towards the end of the network, forming a geographically distributed and heterogeneous set of compute locations, shared and resource-constrained. Such landscape (shared and distributed locations) forces us to design new techniques that can optimize and distribute work among all available locations and, ideally, make compute requirements grow sublinearly with respect to the number of cameras installed. In this paper, we present FoMO (Focus on Moving Objects). This method effectively optimizes multi-camera deployments by preprocessing images for scenes, filtering the empty regions out, and composing regions of interest from multiple cameras into a single image that serves as input for a pre-trained object detection model. Results show that overall system performance can be increased by 8x while accuracy improves 40% as a by-product of the methodology, all using an off-the-shelf pre-trained model with no additional training or fine-tuning.

Let $\pi$ be a set of primes. We show that $\pi$-separable groups have a conjugacy class of $\mathfrak F$-injectors for suitable Fitting classes $\mathfrak F$, which coincide with the usual ones when specializing to soluble groups.

Let $\pi$ be a set of primes. We show that $\pi$-separable groups have a conjugacy class of subgroups which specialize to Carter subgroups, i.e. self-normalizing nilpotent subgroups, or equivalently, nilpotent projectors, when specializing to soluble groups.

The orbit of the outer satellite Alexhelios of (216) Kleopatra is already constrained by adaptive-optics astrometry, obtained with the VLT/SPHERE instrument. However, there is also a preceding occultation event in 1980 attributed to this satellite. Hereinafter, we try to link all observations, spanning 1980--2018. We find the nominal orbit exhibits an unexplained shift by $+60^\circ$ in the true longitude. Using both periodogram analysis and an $\ell = 10$ multipole model suitable for the motion of mutually interacting moons about the irregular body, we confirmed that it is not possible to adjust the respective osculating period $P_2$. Instead, we were forced to use a model with tidal dissipation (and increasing orbital periods) to explain the shift. We also analyzed light curves, spanning 1977--2021, and searched for the expected spin deceleration of Kleopatra. According to our best-fit model, the observed period rate is $\dot P_2 = (1.8\pm 0.1)\cdot 10^{-8}\,{\rm d}\,{\rm d}^{-1}$ and the corresponding time lag $\Delta t_2 = 42\,{\rm s}$ of tides, for the assumed value of the Love number $k_2 = 0.3$. It is the first detection of tidal evolution for moons orbiting 100-km asteroids. The corresponding dissipation factor $Q$ is comparable with other terrestrial bodies, albeit at a higher loading frequency $2|\omega-n|$. We also predict a secular evolution of the inner moon, $\dot P_1 = 5.0\cdot 10^{-8}$, as well as a spin deceleration of Kleopatra, $\dot P_0 = 1.9\cdot 10^{-12}$. In alternative models, with moons captured in the 3:2 mean-motion resonance or more massive moons, the respective values of $\Delta t_2$ are a factor of 2--3 lower. Future astrometric observations by direct imaging or occultations should allow to distinguish between these models, which is important for the internal structure and mechanical properties of (216) Kleopatra., Comment: accepted in A&A

Context. The recent estimates of the 3D shape of the M/Xe-type triple asteroid system (216) Kleopatra indicated a density of 5 g.cm$^{-3}$. Such a high density implies a high metal content and a low porosity which is not easy to reconcile with its peculiar dumbbell shape. Aims. Given the unprecedented angular resolution of the VLT/SPHERE/ZIMPOL camera, we aim to constrain the mass and the shape of Kleopatra with high accuracy, hence its density. Methods. We combined our new VLT/SPHERE observations of Kleopatra recorded in 2017 and 2018 with archival data, as well as lightcurve, occultation, and delay-Doppler images, to derive its 3D shape model using two different algorithms (ADAM, MPCD). Furthermore, an N-body dynamical model allowed us to retrieve the orbital elements of the two moons as explained in the accompanying paper. Results. The shape of Kleopatra is very close to an equilibrium dumbbell figure with two lobes and a thick neck. Its volume equivalent diameter (118.75$\pm$1.40) km and mass (2.97$\pm$0.32) 10$^{18}$ kg imply a bulk density of (3.38$\pm$0.50) g cm$^{-3}$. Such a low density for a supposedly metal-rich body indicates a substantial porosity within the primary. This porous structure along with its near-equilibrium shape is compatible with a formation scenario including a giant impact followed by reaccumulation. Kleopatra's current rotation period and dumbbell shape imply that it is in a critically rotating state. The low effective gravity along the equator of the body, together with the equatorial orbits of the moons and possibly rubble-pile structure, opens the possibility that the moons formed via mass shedding. Conclusions. Kleopatra is a puzzling multiple system due to the unique characteristics of the primary. It deserves particular attention in the future, with the Extremely Large Telescopes and possibly a dedicated space mission., Comment: Accepted for publication in Astronomy and Astrophysics

To interpret adaptive-optics observations of (216) Kleopatra, we need to describe an evolution of multiple moons, orbiting an extremely irregular body and including their mutual interactions. Such orbits are generally non-Keplerian and orbital elements are not constants. Consequently, we use a modified $N$-body integrator, which was significantly extended to include the multipole expansion of the gravitational field up to the order $\ell = 10$. Its convergence was verified against the `brute-force' algorithm. We computed the coefficients $C_{\ell m},S_{\!\ell m}$ for Kleopatra's shape, assuming a~constant bulk density. For solar-system applications, it was also necessary to implement a variable distance and geometry of observations. Our $\chi^2$ metric then accounts for the absolute astrometry, the relative astrometry (2nd moon with respect to 1st), angular velocities, and also silhouettes, constraining the pole orientation. This allowed us to derive the orbital elements of Kleopatra's two moons. Using both archival astrometric data and new VLT/SPHERE observations (ESO LP 199.C-0074), we were able to identify the true periods of the moons, $P_1 = (1.822359\pm0.004156)\,{\rm d}$, $P_2 = (2.745820\pm0.004820)\,{\rm d}$. They orbit very close to the 3:2 mean-motion resonance, but their osculating eccentricities are too small compared to other perturbations (multipole, mutual), so that regular librations of the critical argument are not present. The resulting mass of Kleopatra, $m_1 = (1.49\pm0.16)\cdot10^{-12}\,M_\odot$ or $2.97\cdot10^{18}\,{\rm kg}$, is significantly lower than previously thought. An implication explained in the accompanying paper (Marchis et al.) is that (216) Kleopatra is a critically rotating body., Comment: accepted in A&A

The Rosetta mission provided us with detailed data of the surface of the nucleus of comet 67P/Churyumov-Gerasimenko.In order to better understand the physical processes associated with the comet activity and the surface evolution of its nucleus, we performed a detailed comparative morphometrical analysis of two depressions located in the Ash region. To detect morphological temporal changes, we compared pre- and post-perihelion high-resolution (pixel scale of 0.07-1.75 m) OSIRIS images of the two depressions. We quantified the changes using the dynamic heights and the gravitational slopes calculated from the Digital Terrain Model (DTM) of the studied area using the ArcGIS software before and after perihelion. Our comparative morphometrical analysis allowed us to detect and quantify the temporal changes that occurred in two depressions of the Ash region during the last perihelion passage. We find that the two depressions grew by several meters. The area of the smallest depression (structure I) increased by 90+/-20%, with two preferential growths: one close to the cliff associated with the apparition of new boulders at its foot, and a second one on the opposite side of the cliff. The largest depression (structure II) grew in all directions, increasing in area by 20+/-5%, and no new deposits have been detected. We interpreted these two depression changes as being driven by the sublimation of ices, which explains their global growth and which can also trigger landslides. The deposits associated with depression II reveal a stair-like topography, indicating that they have accumulated during several successive landslides from different perihelion passages. Overall, these observations bring additional evidence of complex active processes and reshaping events occurring on short timescales, such as depression growth and landslides, and on longer timescales, such as cliff retreat.

Judging by popular and generic computer vision challenges, such as the ImageNet or PASCAL VOC, neural networks have proven to be exceptionally accurate in recognition tasks. However, state-of-the-art accuracy often comes at a high computational price, requiring hardware acceleration to achieve real-time performance, while use cases, such as smart cities, require images from fixed cameras to be analyzed in real-time. Due to the amount of network bandwidth these streams would generate, we cannot rely on offloading compute to a centralized cloud. Thus, a distributed edge cloud is expected to process images locally. However, the edge is, by nature, resource-constrained, which puts a limit on the computational complexity that can execute. Yet, there is a need for a meeting point between the edge and accurate real-time video analytics. Specializing lightweight models on a per-camera basis may help but it quickly becomes unfeasible as the number of cameras grows unless the process is automated. In this paper, we present and evaluate COVA (Contextually Optimized Video Analytics), a framework to assist in the automatic specialization of models for video analytics in edge cameras. COVA automatically improves the accuracy of lightweight models through their specialization. Moreover, we discuss and review each step involved in the process to understand the different trade-offs that each one entails. Additionally, we show how the sole assumption of static cameras allows us to make a series of considerations that greatly simplify the scope of the problem. Finally, experiments show that state-of-the-art models, i.e., able to generalize to unseen environments, can be effectively used as teachers to tailor smaller networks to a specific context, boosting accuracy at a constant computational cost. Results show that our COVA can automatically improve accuracy of pre-trained models by an average of 21%.

Convolutions are the core operation of deep learning applications based on Convolutional Neural Networks (CNNs). Current GPU architectures are highly efficient for training and deploying deep CNNs, and hence, these are largely used in production for this purpose. State-of-the-art implementations, however, present a lack of efficiency for some commonly used network configurations. In this paper we propose a GPU-based implementation of the convolution operation for CNN inference that favors coalesced accesses, without requiring prior data transformations. Our experiments demonstrate that our proposal yields notable performance improvements in a range of common CNN forward propagation convolution configurations, with speedups of up to 2.29x with respect to the best implementation of convolution in cuDNN, hence covering a relevant region in currently existing approaches., Comment: This work has been submitted to the Springer for possible publication

The proliferation of machine learning services in the last few years has raised data privacy concerns. Homomorphic encryption (HE) enables inference using encrypted data but it incurs 100x-10,000x memory and runtime overheads. Secure deep neural network (DNN) inference using HE is currently limited by computing and memory resources, with frameworks requiring hundreds of gigabytes of DRAM to evaluate small models. To overcome these limitations, in this paper we explore the feasibility of leveraging hybrid memory systems comprised of DRAM and persistent memory. In particular, we explore the recently-released Intel Optane PMem technology and the Intel HE-Transformer nGraph to run large neural networks such as MobileNetV2 (in its largest variant) and ResNet-50 for the first time in the literature. We present an in-depth analysis of the efficiency of the executions with different hardware and software configurations. Our results conclude that DNN inference using HE incurs on friendly access patterns for this memory configuration, yielding efficient executions., Comment: Manuscript accepted for publication in IEEE Transactions on Computers

Dynamical models of Solar System evolution have suggested that P-/D-type volatile-rich asteroids formed in the outer Solar System and may be genetically related to the Jupiter Trojans, the comets and small KBOs. Indeed, their spectral properties resemble that of anhydrous cometary dust. High-angular-resolution images of P-type asteroid (87) Sylvia with VLT/SPHERE were used to reconstruct its 3D shape, and to study the dynamics of its two satellites. We also model Sylvia's thermal evolution. The shape of Sylvia appears flattened and elongated. We derive a volume-equivalent diameter of 271 +/- 5 km, and a low density of 1378 +/- 45 kg.m-3. The two satellites orbit Sylvia on circular, equatorial orbits. The oblateness of Sylvia should imply a detectable nodal precession which contrasts with the fully-Keplerian dynamics of the satellites. This reveals an inhomogeneous internal structure, suggesting that Sylvia is differentiated. Sylvia's low density and differentiated interior can be explained by partial melting and mass redistribution through water percolation. The outer shell would be composed of material similar to interplanetary dust particles (IDPs) and the core similar to aqueously altered IDPs or carbonaceous chondrite meteorites such as the Tagish Lake meteorite. Numerical simulations of the thermal evolution of Sylvia show that for a body of such size, partial melting was unavoidable due to the decay of long-lived radionuclides. In addition, we show that bodies as small as 130-150 km in diameter should have followed a similar thermal evolution, while smaller objects, such as comets and the KBO Arrokoth, must have remained pristine, in agreement with in situ observations of these bodies. NASA Lucy mission target (617) Patroclus (diameter~140 km) may, however, be differentiated., Comment: Accepted for publication in A&A

The large electrocaloric coupling in PbZrO3 allows using high-speed infrared imaging to visualize antiferroelectric switching dynamics via the associated temperature change. We find that in ceramic samples of homogeneous temperature and thickness, switching is nucleation-limited and fast, with devices responding in the milisecond range. By introducing gradients of thickness, however, it is possible to change the dynamics from nucleation-limited to propagation-limited, whereby a single phase boundary sweeps across the sample like a cold front, at a speed of c.a. 20 cm/s. Additionally, introducing thermostatic temperature differences between two sides of the sample enables the simultaneous generation of a negative electrocaloric effect on one side and a positive one on the other, yielding a Janus-like electrocaloric response.

Context: Until recently, camera networks designed for monitoring fireballs worldwide were not fully automated, implying that in case of a meteorite fall, the recovery campaign was rarely immediate. This was an important limiting factor as the most fragile - hence precious - meteorites must be recovered rapidly to avoid their alteration. Aims: The Fireball Recovery and InterPlanetary Observation Network (FRIPON) scientific project was designed to overcome this limitation. This network comprises a fully automated camera and radio network deployed over a significant fraction of western Europe and a small fraction of Canada. As of today, it consists of 150 cameras and 25 European radio receivers and covers an area of about 1.5 million square kilometers., Comment: 54 pages, 14 figures

Understanding how climate change will affect oceanic fluid transport is crucial for environmental applications and human activities. However, a synoptic characterization of the influence of climate change on mesoscale stirring and transport in the surface ocean is missing. To bridge this gap, we exploit a high-resolution, fully-coupled climate model of the Mediterranean basin using a Network Theory approach. We project significant increases of horizontal stirring and kinetic energies in the next century, likely due to increments of available potential energy. The future evolution of basin-scale transport patterns hint at a rearrangement of the main hydrodynamic provinces, defined as regions of the surface ocean that are well-mixed internally but with minimal cross-flow across their boundaries. This results in increased heterogeneity of province sizes and stronger mixing in their interiors. Our approach can be readily applied to other oceanic regions, providing information for the present and future marine spatial planning.

The conception and design of new musical interfaces is a multidisciplinary area that tightly relates technology and artistic creation. In this paper, the author first exposes some of the questions he has posed himself during more than a decade experience as a performer, composer, interface and software designer, and educator. Finally, he illustrates these topics with some examples of his work., Comment: Proceedings of the International Conference on New Interfaces for Musical Expression, 2001

We have studied the electrocaloric response of the archetypal antiferroelectric PbZrO3 as a function of voltage and temperature in the vicinity of its antiferroelectric-paraelectric phase transition. Large electrocaloric effects of opposite signs, ranging from an electro-cooling of -3.5 K to an electro-heating of +5.5 K, were directly measured with an infrared camera. We show by calorimetric and electromechanical measurements that the large negative electrocaloric effect comes from an endothermic antiferroelectric-ferroelectric switching, in contrast to dipole destabilization of the antiparallel lattice, previously proposed as an explanation for the negative electrocaloric effect of antiferroelectrics., Comment: Article (17 pages) and supplemental material (12 pages) present in .pdf file

Asteroid (31) Euphrosyne is one of the biggest objects in the asteroid main belt and the Euphrosyne family occupies a highly inclined region in the outer main belt and contains a remarkably large number of members, which is interpreted as an outcome of a disruptive cratering event. The goals of this adaptive-optics imaging study were threefold: to characterize the shape of Euphrosyne, to constrain its density, and to search for the large craters that may be associated with the family formation event. We obtained disk-resolved images of Euphrosyne using SPHERE/ZIMPOL at ESO's 8.2-m VLT as part of our large program (ID: 199.C-0074, PI: Vernazza). We reconstructed its 3D-shape using the adam shape modeling algorithm based on the SPHERE images and the available lightcurves of this asteroid. We analyzed the dynamics of the satellite with the genoid meta-heuristic algorithm. Finally, we studied the shape of Euphrosyne using hydrostatic equilibrium models. Our SPHERE observations show that Euphrosyne has a nearly spherical shape with the sphericity index of 0.9888 and its surface lacks large impact craters. Euphrosyne's diameter is 268+/-6 km, making it one of the top 10 largest main belt asteroids. We detected a satellite of Euphrosyne -- S/2019 (31) 1-- that is about 4 km across, on an circular orbit. The mass determined from the orbit of the satellite together with the volume computed from the shape model imply a density of 1665+/-242 kg/m^3, suggesting that Euphrosyne probably contain a large fraction of water ice in its interior. We find that the spherical shape of Euphrosyne is a result of the reaccumulation process following the impact, as in the case of (10) Hygiea. However, our shape analysis reveals that, contrary to Hygiea, the axis ratios of Euphrosyne significantly differ from the ones suggested by fluid hydrostatic equilibrium following reaccumulation., Comment: 12 pages, 6 figures, accepted to be published in A&A

Massive exploitation of next-generation sequencing technologies requires dealing with both: huge amounts of data and complex bioinformatics pipelines. Computing architectures have evolved to deal with these problems, enabling approaches that were unfeasible years ago: accelerators and Non-Volatile Memories (NVM) are becoming widely used to enhance the most demanding workloads. However, bioinformatics workloads are usually part of bigger pipelines with different and dynamic needs in terms of resources. The introduction of Software Defined Infrastructures (SDI) for data centers provides roots to dramatically increase the efficiency in the management of infrastructures. SDI enables new ways to structure hardware resources through disaggregation, and provides new hardware composability and sharing mechanisms to deploy workloads in more flexible ways. In this paper we study a state-of-the-art genomics application, SMUFIN, aiming to address the challenges of future HPC facilities., Comment: Partially funded by European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement No 639595) - HiEST Project Published Euro-Par 2018: Euro-Par 2018: Parallel Processing Workshops pp 613-625

This study presents methane-air counterflow simulations, in computationally efficient similar form, allowing combustible mixtures to flow from one or both directions in order to learn more about multi-branched propagating flame structures (e.g., a triple flame). These structures with both premixed and non-premixed flames are commonly seen in more practical combustion analyses. A range of realistic mass mixture fractions and asymmetric chemical rate laws are examined while avoiding the commonly forced unreal symmetric behavior with one-step second-order kinetics. Moreover, a survey of critical parameters is performed varying pressure and normal strain rate to define the flame structure and detect different characters. Three flames can co-exist if the strain rate is low enough and the pressure is high enough. However, at higher strain rate and/or lower pressure, only one or two flames might be obtained. Negative regions of heat release rate are observed and linked to potential endothermic reactions. With a rich premixed mixture at low strain rates and pressures, high exothermic reactions producing CO$_2$ and H$_2$O, and consuming CO and H$_2$ causes a heat-release-rate peak. Unexpected character of the lean and rich premixed flames is observed, leading to the conclusion that these flames are diffusion-controlled.

The considerable impact of Convolutional Neural Networks on many Artificial Intelligence tasks has led to the development of various high performance algorithms for the convolution operator present in this type of networks. One of these approaches leverages the \imcol transform followed by a general matrix multiplication (GEMM) in order to take advantage of the highly optimized realizations of the GEMM kernel in many linear algebra libraries. The main problems of this approach are 1) the large memory workspace required to host the intermediate matrices generated by the IM2COL transform; and 2) the time to perform the IM2COL transform, which is not negligible for complex neural networks. This paper presents a portable high performance convolution algorithm based on the BLIS realization of the GEMM kernel that avoids the use of the intermediate memory by taking advantage of the BLIS structure. In addition, the proposed algorithm eliminates the cost of the explicit IM2COL transform, while maintaining the portability and performance of the underlying realization of GEMM in BLIS.