Search

Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 3,000 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai., Comment: 25 pages, 6 figures

Explicit classical states achieving maximal $f$-divergence are given, allowing for a simple proof of Matsumoto's Theorem, and the systematic extension of any inequality between classical $f$-divergences to quantum $f$-divergences. Our methodology is particularly simple as it does not require any elaborate matrix analysis machinery but only basic linear algebra. It is also effective, as illustrated by two examples improving existing bounds: (i)~an improved quantum Pinsker inequality is derived between $\chi^2$ and trace norm, and leveraged to improve a bound in decoherence theory; (ii)~a new reverse quantum Pinsker inequality is derived for any quantum $f$-divergence, and compared to previous (Audenaert-Eisert and Hirche-Tomamichel) bounds.

Accurate positioning is crucial in the construction industry, where labor shortages highlight the need for automation. Robotic systems with long kinematic chains are required to reach complex workspaces, including floors, walls, and ceilings. These requirements significantly impact positioning accuracy due to effects such as deflection and backlash in various parts along the kinematic chain. In this work, we introduce a novel approach that integrates deflection and backlash compensation models with high-accuracy accelerometers, significantly enhancing position accuracy. Our method employs a modular framework based on a factor graph formulation to estimate the state of the kinematic chain, leveraging acceleration measurements to inform the model. Extensive testing on publicly released datasets, reflecting real-world construction disturbances, demonstrates the advantages of our approach. The proposed method reduces the $95\%$ error threshold in the xy-plane by $50\%$ compared to the state-of-the-art Virtual Joint Method, and by $31\%$ when incorporating base tilt compensation., Comment: 8 pages, 7 figures, Accepted on November 2024 at IEEE Robotics and Automation Letters

Given a family $\mathcal{H}$ of graphs, we say that a graph $G$ is $\mathcal{H}$-free if no induced subgraph of $G$ is isomorphic to a member of $\mathcal{H}$. Let $S_{t,t,t}$ be the graph obtained from $K_{1,3}$ by subdividing each edge $t-1$ times, and let $W_{t\times t}$ be the $t$-by-$t$ hexagonal grid. Let $\mathcal{L}_t$ be the family of all graphs $G$ such that $G$ is the line graph of some subdivision of $W_{t \times t}$. We prove that for every positive integer $t$ there exists $c(t)$ such that every $\mathcal{L}_t \cup \{S_{t,t,t}, K_{t,t}\}$-free $n$-vertex graph admits a tree decomposition in which the maximum size of an independent set in each bag is at most $c(t)\log^4n$. This is a variant of a conjecture of Dallard, Krnc, Kwon, Milani\v{c}, Munaro, \v{S}torgel, and Wiederrecht from 2024. This implies that the Maximum Weight Independent Set problem, as well as many other natural algorithmic problems, that are known to be NP-hard in general, can be solved in quasi-polynomial time if the input graph is $\mathcal{L}_t \cup \{S_{t,t,t},K_{t,t}\}$-free. As part of our proof, we show that for every positive integer $t$ there exists an integer $d$ such that every $\mathcal{L}_t \cup \{S_{t,t,t}\}$-free graph admits a balanced separator that is contained in the neighborhood of at most $d$ vertices.

We study connection points on the double regular $n$-gon translation surface, for $n \geq 7$ odd and its staircase model. For $n \neq 9$, we provide a large family of points with coordinates in the trace field that are not connection points. This family includes the central points, and for $n=7$ we conjecture that all the remaining points are connection points. Further, in the case where $n \geq 7$ is a prime number, we provide a constructive proof by exhibiting an explicit separatrix passing through a central point that does not extend to a saddle connection., Comment: 25 pages, 7 figures, comments welcome

Quantum coherence in bosonic systems is a fundamental resource for quantum technology applications. In this work, we introduce a framework for analyzing coherence in the Fock-state basis, utilizing context-dependent certification to reveal the quantum non-Gaussian nature of the tested coherence. Rather than relying on global coherence measures, our approach targets specific aspects of coherence, enabling a tailored hierarchical classification. We derive and compare two distinct hierarchies, each representing a different context for coherence in bosonic systems. Motivated by current advancements in optical quantum state engineering, we assess the feasibility and depth of these hierarchies under conditions of loss and thermal noise. The methodology introduced here is versatile and can be extended to multi-state and multi-mode coherences, making it adaptable to a wide range of experimental scenarios and platforms.

Non-Gaussianity, a distinctive characteristic of bosonic quantum states, is pivotal in advancing quantum networks, fault-tolerant quantum computing, and high-precision metrology. Verifying the quantum nature of a state, particularly its non-Gaussian features, is essential for ensuring the reliability and performance of these technologies. However, the specific properties required for each application demand tailored validation thresholds. Here, we introduce a hierarchical framework comprising absolute, relative, and qubit-specific thresholds to assess the non-Gaussianity of local coherences. We illustrate this framework using heralded optical non-Gaussian states with the highest purities available in optical platforms. This comprehensive framework presents the first detailed evaluation of number state coherences and can be extended to a wide range of bosonic states.

Improving the efficiency of photothermal (PT) therapies and photoacoustic (PA) imaging at the microscopic scale hinges on developing multiphoton-absorbing photothermal molecules or contrast agents that operate in the near-infrared (NIR) range. These advanced agents or molecules will enable excitation with NIR lasers, in an improved transparency range for biological tissues, while enabling minute, highly localized spatial control of the excitation area. However, progress in this field requires innovative experimental techniques to characterize photothermal and photoacoustic effects under multiphoton excitation. In this article, we showcase a study of a model organometallic molecular compound excited via two-photon absorption (2PA) using femtosecond laser pulses. Based on a time-domain Brillouin scattering technique, well adapted for investigating ultrafast nonlinear optical absorption processes in ultrathin films on substrates, we determine the effective nonlinear absorption coefficients of the compound directly linked to PT/PA. Our findings provide a practical approach for exploring and optimizing nonlinear PT/PA absorbers and contrast agents.

Robust surfaces capable of reducing flow drag, controlling heat and mass transfer, and resisting fouling in fluid flows are important for various applications. In this context, textured surfaces impregnated with a liquid lubricant show promise due to their ability to sustain a liquid-liquid layer that induces slippage. However, theoretical and numerical studies suggest that the slippage can be compromised by surfactants in the overlying fluid, which contaminate the liquid-liquid interface and generate Marangoni stresses. In this study, we use Doppler-optical coherence tomography, an interferometric imaging technique, combined with numerical simulations to investigate how surfactants influence the slip length of lubricant-infused surfaces with longitudinal grooves in a laminar flow. We introduce surfactants by adding tracer particles (milk) to the working fluid (water). Local measurements of slip length at the liquid-liquid interface are significantly smaller than theoretical predictions for clean interfaces (Sch\"onecker & Hardt 2013). In contrast, measurements are in good agreement with numerical simulations of fully immobilized interfaces, indicating that milk particles adsorbed at the interface are responsible for the reduction in slippage. This work provides the first experimental evidence that liquid-liquid interfaces within textured surfaces can become immobilized in the presence of surfactants and flow., Comment: 28 pages, 18 figures

Muon-spin-rotation ($\mu$SR) experiments and the observation of a spontaneous Nernst effect indicate time-reversal symmetry breaking (BTRS) at $T_{\rm c}^{\rm Z2}$ above the superconducting transition temperature $T_{\rm c}$ in Ba$_{1-x}$K$_x$Fe$_2$As$_2$, with $x\approx0.8$. Further studies have pointed out that BTRS is caused by the formation of a new state of matter associated with the condensation of pairs of electron pairs. Despite exhibiting multiple unconventional effects that warrant further investigation, the electronic spectral properties of this electron quadrupling state remain largely unexplored. Here, we present detailed $^{75}$As nuclear magnetic resonance (NMR) measurements of Ba$_{1-x}$K$_x$Fe$_2$As$_2$, with $x = 0.77$, which has $T_{\rm c}^{\rm Z2}$ > $T_{\rm c}$ according to measurements of the spontaneous Nernst effect. The NMR data obtained in this work provide the first direct electronic spectral characteristics of the electron quadrupling state by indicating that it evolves from a pseudogap that sets in at $T^*$ well above $T_{\rm c}^{\rm Z2}$. This pseudogap behavior is consistent with $\mu$SR Knight-shift, specific-heat, and transport data indicating the formation of a bound state of electrons. According to a theory of electron quadrupling condensates, such bound-state formations should precede the onset of BTRS correlations between pairs of electron pairs. The second important insight from NMR data is the absence of spin-related magnetism. The temperature dependence of the spin-lattice relaxation rate $1/T_1T$ and the evolution of the NMR linewidth prove the absence of a magnetic transition at $T_{\rm c}^{\rm Z2}$ and rule out even a proximity to some magnetic instability. This indicates that the spontaneous magnetic fields detected in this compound are not caused by spin magnetism but are associated with persistent real-space currents., Comment: 22 pages, 18 figures

We explore the ability of two LLMs -- GPT-4o and Claude Sonnet 3.5 -- to transcribe historical handwritten documents in a tabular format and compare their performance to traditional OCR/HTR systems: EasyOCR, Keras, Pytesseract, and TrOCR. Considering the tabular form of the data, two types of experiments are executed: one where the images are split line by line and the other where the entire scan is used as input. Based on CER and BLEU, we demonstrate that LLMs outperform the conventional OCR/HTR methods. Moreover, we also compare the evaluated CER and BLEU scores to human evaluations to better judge the outputs of whole-scan experiments and understand influential factors for CER and BLEU. Combining judgments from all the evaluation metrics, we conclude that two-shot GPT-4o for line-by-line images and two-shot Claude Sonnet 3.5 for whole-scan images yield the transcriptions of the historical records most similar to the ground truth., Comment: 15 pages, 7 figures

This work is motivated by analyses of longitudinal data collected from participants in the Quebec Longitudinal Study of Child Development (QLSCD) and the Quebec Newborn Twin Study (QNTS) to identify important genetic predictors for emotional and behavioral difficulties in childhood and adolescence. We propose a lasso penalized mixed model for continuous and binary longitudinal traits that allows the inclusion of multiple random effects to account for random individual effects not attributable to the genetic similarity between individuals. Through simulation studies, we show that replacing the estimated genetic relatedness matrix (GRM) by a sparse matrix introduces bias in the variance components estimates, but that the obtained computational gain is major while the impact on the performance of the penalized model to retrieve important predictors is negligible. We compare the performance of the proposed penalized mixed model to a standard lasso and to a univariate mixed model association test and show that the proposed model always identifies causal predictors with greater precision. Finally, we show an application of the proposed methodology to predict three externalizing behavorial scores in the combined QLSCD and QNTS longitudinal cohorts., Comment: 46 pages, 9 figures

1) Objectives: This technical report presents a synthetic summary and the principal outcomes of the project R-Vessel-X ("Robust vascular network extraction and understanding within hepatic biomedical images") funded by the French Agence Nationale de la Recherche, and developed between 2019 and 2023. 2) Material and methods: We used datasets and tools publicly available such as IRCAD, Bullitt or VascuSynth toobtain real or synthetic angiographic images. The main contributions lie in the field of 3D angiographic image analysis: filtering, segmentation, modeling and simulation, with a specific focus on the liver. 3) Results: We paid a particular attention to open-source software diffusion of the developed methods, by means of 3D Slicer plugins for the liver anatomy segmentation (SlicerRVXLiverSegmentation) and vesselness filtering (Slicer-RVXVesselnessFilters), and an online demo for the generation of synthetic and realistic vessels in 2D and 3D (OpenCCO). 4) Conclusion: The R-Vessel-X project provided extensive research outcomes, covering various topics related to 3D angiographic image analysis, such as filtering, segmentation, modeling and simulation. We also developed open-source and free softwares so that the research communities in biomedical engineering can use these results in their future research., Comment: Innovation and Research in BioMedical engineering, In press

To detect anomalies with precision and without prior knowledge in time series, is it better to build a detector from the initial temporal representation, or to compute a new (tabular) representation using an existing automatic variable construction library? In this article, we address this question by conducting an in-depth experimental study for two popular detectors (Isolation Forest and Local Outlier Factor). The obtained results, for 5 different datasets, show that the new representation, computed using the tsfresh library, allows Isolation Forest to significantly improve its performance., Comment: 7

Neutrino transport in compact objects is an inherently challenging multi-dimensional problem. This difficulty is compounded if one includes flavor transformation -- an intrinsically quantum phenomenon requiring one to follow the coherence between flavors and thus necessitating the introduction of complex numbers. To reduce the computational burden, simulations of compact objects that include neutrino transport often make use of momentum-angle-integrated moments (the lowest order ones being commonly referred to as the energy density and flux) and these quantities can be generalized to include neutrino flavor, i.e., they become quantum moments. Numerous finite-volume approaches to solving the moment evolution equations for classical neutrino transport have been developed based on solving a Riemann problem at cell interfaces. In this paper we describe our generalization of a Riemann solver for quantum moments, specifically decomposing complex numbers in terms of a (signed) magnitude and phase instead of real and imaginary parts. We then test our new algorithm in numerous cases showing a neutrino fast flavor instability, varying from toy models with analytic solutions to snapshots from neutron star merger simulations. Compared to previous algorithms for neutrino transport with flavor mixing, we find uniformly smaller growth rates of the flavor transformation along with concomitantly larger length-scales, and that the results are a better match with the growth rates seen from multi-angle codes., Comment: 31 pages, 3 appendices, 25 figures

Artificial intelligence has been integrated into nearly every aspect of daily life, powering applications from object detection with computer vision to large language models for writing emails and compact models in smart homes. These machine learning models cater to individual users but are often detached from them, as they are typically stored and processed in centralized data centers. This centralized approach raises privacy concerns, incurs high infrastructure costs, and struggles with personalization. Federated and fully decentralized learning methods have been proposed to address these issues, but they still depend on centralized servers or face slow convergence due to communication constraints. To overcome these challenges, we propose ML Mule, a approach that utilizes individual mobile devices as 'Mules' to train and transport model snapshots as they move through physical spaces, sharing these models with the physical 'Spaces' they inhabit. This method implicitly forms affinity groups among devices associated with users who share particular spaces, enabling collaborative model evolution, and protecting users' privacy. Our approach addresses several major shortcomings of traditional, federated, and fully decentralized learning systems. The proposed framework represents a new class of machine learning methods that are more robust, distributed, and personalized, bringing the field closer to realizing the original vision of intelligent, adaptive, and genuinely context-aware smart environments. The results show that ML Mule converges faster and achieves higher model accuracy compared to other existing methods.

Accurate sensor calibration is crucial for autonomous systems, yet its uncertainty quantification remains underexplored. We present the first approach to integrate uncertainty awareness into online extrinsic calibration, combining Monte Carlo Dropout with Conformal Prediction to generate prediction intervals with a guaranteed level of coverage. Our method proposes a framework to enhance existing calibration models with uncertainty quantification, compatible with various network architectures. Validated on KITTI (RGB Camera-LiDAR) and DSEC (Event Camera-LiDAR) datasets, we demonstrate effectiveness across different visual sensor types, measuring performance with adapted metrics to evaluate the efficiency and reliability of the intervals. By providing calibration parameters with quantifiable confidence measures, we offer insights into the reliability of calibration estimates, which can greatly improve the robustness of sensor fusion in dynamic environments and usefully serve the Computer Vision community., Comment: Accepted for publication at WACV 2025

Neural network verification is a new and rapidly developing field of research. So far, the main priority has been establishing efficient verification algorithms and tools, while proper support from the programming language perspective has been considered secondary or unimportant. Yet, there is mounting evidence that insights from the programming language community may make a difference in the future development of this domain. In this paper, we formulate neural network verification challenges as programming language challenges and suggest possible future solutions., Comment: Accepted at ESOP 2025, European Symposium on Programming Languages

We show that it is NP-hard to distinguish graphs of linear mim-width at most 1211 from graphs of sim-width at least 1216. This implies that Mim-Width, Sim-Width, One-Sided Mim-Width, and their linear counterparts are all paraNP-complete, i.e., NP-complete to compute even when upper bounded by a constant., Comment: 27 pages, 9 figures

The search for resonant mass bumps in invariant-mass distributions remains a cornerstone strategy for uncovering Beyond the Standard Model (BSM) physics at the Large Hadron Collider (LHC). Traditional methods often rely on predefined functional forms and exhaustive computational and human resources, limiting the scope of tested final states and selections. This work presents BumpNet, a machine learning-based approach leveraging advanced neural network architectures to generalize and enhance the Data-Directed Paradigm (DDP) for resonance searches. Trained on a diverse dataset of smoothly-falling analytical functions and realistic simulated data, BumpNet efficiently predicts statistical significance distributions across varying histogram configurations, including those derived from LHC-like conditions. The network's performance is validated against idealized likelihood ratio-based tests, showing minimal bias and strong sensitivity in detecting mass bumps across a range of scenarios. Additionally, BumpNet's application to realistic BSM scenarios highlights its capability to identify subtle signals while managing the look-elsewhere effect. These results underscore BumpNet's potential to expand the reach of resonance searches, paving the way for more comprehensive explorations of LHC data in future analyses.

The shape of human brain is complex and highly variable, with interactions between brain size, cortical folding, and age well-documented in the literature. However, few studies have explored how global brain size influences geometric features of the cortical surface derived from anatomical MRI. In this work, we focus on sulcal depth, an imaging phenotype that has gained significant attention in both basic research and clinical applications. We make key contributions to the field by: 1) providing the first quantitative analysis of how brain size affects sulcal depth measurements; 2) introducing a novel, scale-invariant method for sulcal depth estimation based on an original formalization of the problem; 3) presenting a validation framework and sharing our code and benchmark data with the community; and 4) demonstrating the biological relevance of our new sulcal depth measure using a large sample of 1,987 subjects spanning the developmental period from 26 weeks post-conception to adulthood., Comment: GA and JL contributed equally to this work

We study the dynamics of a zero-temperature overdamped tracer in a bath of Brownian particles. As the bath density is increased, the passive tracer transitions from an effectively active dynamics, characterized by boundary accumulation and ratchet currents, to a bona-fide equilibrium regime. To account for this, we eliminate the bath degrees of freedom under the assumption of linear coupling to the tracer and show convergence to equilibrium in the large density limit. We then develop a perturbation theory to characterize the tracer's departure from equilibrium at large but finite bath densities, revealing an intermediate time-reversible yet non-Boltzmann regime, followed by a fully irreversible one. Finally, we show that when the bath particles are connected as a lattice, mimicking a gel, the cold tracer drives the entire bath out of equilibrium, leading to a long-ranged suppression of bath fluctuations., Comment: 6 pages, 4 figures

We investigate the formation of helical multifilament bundles and the torque required to achieve them as a function of applied twist. Hyperelastic filaments with circular cross sections are mounted parallel in a uniform circle onto end-clamps that can move along the twist axis depending on the applied axial load. With increasing twist, the filaments describe a hyperbolic hyperboloid surface before coming into contact in a circle, and then packing in a tight helical bundle in the center with increasing twist. While the bundle appears ordered for sufficiently small number of filaments, they are disordered for large enough number of filaments and applied twist. We reveal with x-ray tomography, that the packing of the filaments becomes disordered following a radial-instability which leads to a decrease in bundle radius, and migration of filaments relative to each other in the bundle. Nonetheless, the helical angle of the filaments in the bundle are found to be essentially constant, resulting in inclination angles which increase with distance from axis of rotation. We develop energy minimization analysis to capture the observed variations in bundle length and torque as a function of number of filaments considering the neo-Hookean nature of the filaments. We show that the bundle geometry and the applied load can be used to describe the non-linear torque profile measured as a function of twist angle.

We present the official release of the EFfective Field theORy surrogaTe (Effort), a novel and efficient emulator designed for the Effective Field Theory of Large-Scale Structure (EFTofLSS). This tool combines state-of-the-art numerical methods and clever preprocessing strategies to achieve exceptional computational performance without sacrificing accuracy. To validate the emulator reliability, we compare Bayesian posteriors sampled using Effort via Hamiltonian MonteCarlo methods to the ones sampled using the widely-used pybird code, via the Metropolis-Hastings sampler. On a large-volume set of simulations, and on the BOSS dataset, the comparison confirms excellent agreement, with deviations compatible with MonteCarlo noise. Looking ahead, Effort is poised to analyze next-generation cosmological datasets and to support joint analyses with complementary tools., Comment: 25 pages, 5 figure. Comments are welcome!

An excellent estimate of the lensing signal is expected from the availability of deep and high-resolution polarization data in the near future. This is most important to allow for efficient delensing, needed to detect the primordial B-mode power and with it the famous tensor-to-scalar ratio. Here we discuss in a joint manner estimators of the rotation of polarization, of the second order lensing field rotation, and standard gradient lensing reconstruction. All are most efficient when able to probe the EB power created locally, have comparable reconstruction noise in this regime, and can benefit substantially from delensing. We discuss several ongoing and planned CMB experiments. We determine their noise for lensing field rotation and polarization rotation and discuss their prospects for measuring these effects. There is an on-going controversy on whether the lensing field rotation also rotates the polarization -- if so this will be observed at high significance soon with already on going observations of the South Pole Telescope, SPT-3G, in cross-correlation with tracers of large scale structure, as we show in this paper., Comment: 20 pages, 8 figures, some appendices

Let M be a closed 3-manifold and S(M) the skein module of M at some odd root of unity. Using the Frobenius morphism, we can see S(M) as the space of global sections of a coherent sheaf over the SL2 character scheme of M. We prove that when the character scheme is reduced, this sheaf is a line bundle.

Ekman pumping is a phenomenon induced by no-slip boundary conditions in rotating fluids. In the context of Rayleigh-B\'enard convection, Ekman pumping causes a significant change in the linear stability of the system compared to when it is not present (that is, stress-free). Motivated by numerical solutions to the marginal stability problem of the incompressible Navier-Stokes (iNSE) system, we seek analytical asymptotic solutions which describe the departure of the no-slip solution from the stress-free. The substitution of normal modes into a reduced asymptotic model yields a linear system for which we explore analytical solutions for various scalings of wavenumber. We find very good agreement between the analytical asymptotic solutions and the numerical solutions to the iNSE linear stability problem with no-slip boundary conditions.

When submerged in a chiral active bath, a passive object becomes a spinning ratchet imbued with odd transport properties. In the adiabatic limit of a massive object, we derive the most general Langevin dynamics for a rigid body in a chiral active bath, with odd diffusion and odd mobility connected by an Einstein relation, and numerically explore the breakdown of these predictions beyond the adiabatic limit. We show that the irreversibility of a massive object increases as its symmetry decreases: a disk exhibits an effective equilibrium dynamics, while a rod admits distinct translational and rotational temperatures, and a wedge is fully out of equilibrium. Conversely, this departure from equilibrium can be read in the universal far-field currents and density modulations of the bath, which we measure numerically and derive analytically.

Triple Photon Generation (TPG) is a third-order nonlinear optical interaction in which a photon, i.e. the pump, splits into three lower energy photons, i.e. modes 1, 2 and 3. The triplets possess different quantum signatures from those of photon pairs, with a strong interest in quantum information. In the present study, we performed the first experimental demonstration of TPG stimulated over one mode of the triplet, mode 1, the previous work on TPG concerning stimulation over two modes,2 and 3. The nonlinear medium is a KTiOPO4 crystal pumped in the picosecond regime (15 ps, 10 Hz) at a pump wavelength of 532 nm. The stimulation beam is emitted by a tunable optical parametric generator: the phase-matching was found at a stimulation wavelength of 1491 nm, the other two modes of the triplet being both at 1654 nm in orthogonal polarizations. Using superconducting nanowires single photon detectors, the measurement of the polarizations and wavelength signatures of the two generated modes are in full agreement with calculations. It has been possible to generate a total number of photons per pulse on modes 2 and 3 up to 2x10-4, which corresponds to the generation of 10-4 triplets per pulse, or 10-5 triplets per second since the repetition rate is equal to 10 Hz. We interpreted these results in the framework of a model we developed on the basis of the nonlinear momentum operator in the Heisenberg representation under the undepleted pump and stimulation approximation., Comment: 6 pages, 5 figures

With today's multi-petawatt lasers, testing quantum electrodynamics (QED) in the strong field regime, where the electric field exceeds the Schwinger critical field in the rest frame of an electron, becomes within reach. Inverse Compton scattering of an intense laser pulse off a high-energy electron beam is the mainstream approach, resulting in the emission of high-energy photons that can decay into Breit-Wheeler electron-positron pairs. Here, we demonstrate experimentally that very high energy photons can be generated in a self-aligned single-laser Compton scattering setup, combining a laser-plasma accelerator and a plasma mirror. Reaching up to the GeV scale, photon emission via nonlinear Compton scattering exhibits a nonclassical scaling in the experiment that is consistent with electric fields reaching up to a fraction $\chi\simeq0.3$ of the Schwinger field in the electron rest frame. These foolproof collisions guaranteed by automatic laser-electron overlap provide a new approach for precise investigations of strong-field QED processes.

Contrastive learning has proven effective in self-supervised musical representation learning, particularly for Music Information Retrieval (MIR) tasks. However, reliance on augmentation chains for contrastive view generation and the resulting learnt invariances pose challenges when different downstream tasks require sensitivity to certain musical attributes. To address this, we propose the Leave One EquiVariant (LOEV) framework, which introduces a flexible, task-adaptive approach compared to previous work by selectively preserving information about specific augmentations, allowing the model to maintain task-relevant equivariances. We demonstrate that LOEV alleviates information loss related to learned invariances, improving performance on augmentation related tasks and retrieval without sacrificing general representation quality. Furthermore, we introduce a variant of LOEV, LOEV++, which builds a disentangled latent space by design in a self-supervised manner, and enables targeted retrieval based on augmentation related attributes.

Since the discovery of heavy-fermion superconductivity in CeCu$_{2}$Si$_{2}$, the material has attracted great interest particularly with regard to the nature of the superconducting pairing and its mechanism. Consequently, it is essential to better understand the electronic Fermi surface topology and its role in strong antiferromagnetic fluctuations. The standard density functional theory method is insufficient to model the interplay of strong onsite Coulomb repulsion in localized 4{\it f}-electrons and their hybridization with itinerant ligand-orbital electrons. We have performed electronic ground state calculations on CeCu$_{2}$Si$_{2}$ using the Gutzwiller wavefunction approximation. The Gutzwiller approximation captures the quasiparticle band renormalization from the strong onsite Coulomb repulsion. We have performed an analysis of this effect on the electronic structure and the Fermi surface topology by varying the interaction strength and taking into account the crystal-field splitting. Using the de Haas van Alphen effect, the extremal Fermi surface cross-sectional areas were calculated to quantify the effects of quasiparticle mass renormalization on the Fermi surface. Our results confirm two Fermi surface sheets corresponding to the heavy (488m$_{e}$) and light (4.35m$_{e}$) quasiparticles, which is in close agreement with experimental measurements as well as the renormalized band method., Comment: 10 pages, 3 figures, 2 tables

We survey various constructions of finite dimensional projective representations of mapping class groups derived from stated skein algebras.

Stellar surface heterogeneities, such as spots and faculae, often contaminate exoplanet transit spectra, hindering precise atmospheric characterization. We demonstrate a novel, epoch-based, model-independent method to mitigate stellar contamination, applicable to multi-planet systems with at least one airless planet. We apply this method using quasi-simultaneous transits of TRAPPIST-1 b and TRAPPIST-1 c observed on July 9, 2024, with JWST NIRSpec PRISM. These two planets, with nearly identical radii and impact parameters, are likely either bare rocks or possess thin, low-pressure atmospheres, making them ideal candidates for this technique, as variations in their transit spectra would be primarily attributed to stellar activity. Our observations reveal their transit spectra exhibit consistent features, indicating similar levels of stellar contamination. We use TRAPPIST-1 b to correct the transit spectrum of TRAPPIST-1 c, achieving a 2.5x reduction in stellar contamination at shorter wavelengths. At longer wavelengths, lower SNR prevents clear detection of contamination or full assessment of mitigation. Still, out-of-transit analysis reveals variations across the spectrum, suggesting contamination extends into the longer wavelengths. Based on the success of the correction at shorter wavelengths, we argue that contamination is also reduced at longer wavelengths to a similar extent. This shifts the challenge of detecting atmospheric features to a predominantly white noise issue, which can be addressed by stacking observations. This method enables epoch-specific stellar contamination corrections, allowing co-addition of planetary spectra for reliable searches of secondary atmospheres with signals of 60-250 ppm. Additionally, we identify small-scale cold (2000 K) and warm (2600 K) regions almost uniformly distributed on TRAPPIST-1, with overall covering fractions varying by 0.1% per hour., Comment: Resubmitted to ApJL after addressing reviewer comments

Lactate is crucial in the brain, as it is involved in neuronal activity and memory formation. This is thought to be related to the astrocyte-neuron lactate shuttle hypothesis (ANLS), which has been debated for 30 years, not least because it is difficult to measure lactate compartmentation and exchange in vivo. In particular, while ANLS requires transfer of lactate via the extracellular space, intracellular/extracellular exchange rate remains unknown. In this work, we propose to use time-dependent diffusion MRS to assess lactate exchange and lactate compartmentation in mouse brain gray matter in vivo. First, by comparing lactate time-dependent diffusivity and kurtosis with those of water and purely intracellular metabolites (which exhibit "fast" and "no" exchange signatures, respectively), we estimate that lactate exchange is slow (i.e. on the order of hundreds of milliseconds). Then, using biophysical models neglecting exchange, we estimate extracellular, neuronal and astrocytic lactate fractions compatible with the ANLS hypothesis.

We study the dynamics of multipulse solutions in mode-locked lasers in presence of time-delayed feedback stemming, e.g., from reflections upon optical elements, and carrier dynamics. We demonstrate that the dynamics of such a high dimensional problem can be successfully described by some effective equations of motion for the pulses' phases and positions. Analyzing the reduced vector field permits disclosing a highly complex dynamics where coherent and incoherent interactions compete. The latter lead to regimes in which pulses can be equidistant or non-equidistant and also have different phase relations. Multi-stability between regimes is also observed as well as emerging limit cycles and global heteroclinic bifurcations in phase space.

Supermassive black holes are prevalent at the centers of massive galaxies, and their masses scale with galaxy properties, increasing evidence suggesting that these trends continue to low stellar masses. Seeds are needed for supermassive black holes, especially at the highest redshifts explored by the James Webb Space Telescope. We study the hierarchical merging of galaxies via cosmological merger trees and argue that the seeds of supermassive black holes formed in nuclear star clusters via stellar black hole mergers at early epochs. Observable tracers include intermediate-mass black holes, nuclear star clusters, and early gas accretion in host dwarf galaxies, along with a potentially detectable stochastic gravitational wave background, ejection of intermediate and supermassive black holes, and consequences of a significant population of tidal disruption events and extreme-mass ratio inspirals., Comment: 20 pages, 12 figures

Consider a graph $G$ with a long path $P$. When is it the case that $G$ also contains a long induced path? This question has been investigated in general as well as within a number of different graph classes since the 80s. We have recently observed in a companion paper (Long induced paths in sparse graphs and graphs with forbidden patterns, arXiv:2411.08685, 2024) that most existing results can recovered in a simple way by considering forbidden ordered patterns of edges along the path $P$. In particular we proved that if we forbid some fixed ordered matching along a path of order $n$ in a graph $G$, then $G$ must contain an induced path of order $(\log n)^{\Omega(1)}$. Moreover, we completely characterized the forbidden ordered patterns forcing the existence of an induced path of polynomial size. The purpose of the present paper is to completely characterize the ordered patterns $H$ such that forbidding $H$ along a path $P$ of order $n$ implies the existence of an induced path of order $(\log n)^{\Omega(1)}$. These patterns are star forests with some specific ordering, which we called constellations. As a direct consequence of our result, we show that if a graph $G$ has a path of length $n$ and does not contain $K_t$ as a topological minor, then $G$ contains an induced path of order $(\log n)^{\Omega(1/t \log^2 t)}$. The previously best known bound was $(\log n)^{f(t)}$ for some unspecified function $f$ depending on the Topological Minor Structure Theorem of Grohe and Marx (2015)., Comment: Some material from arXiv:2304.09679 has been reused (such as definitions and remarks), so text overlap is to be expected

We extend the scope of analytic combinatorics to classes containing objects that have irrational sizes. The generating function for such a class is a power series that admits irrational exponents (which we call a Ribenboim series). A transformation then yields a generalised Dirichlet series from which the asymptotics of the coefficients can be extracted by singularity analysis using an appropriate Tauberian theorem. In practice, the asymptotics can often be determined directly from the original generating function. We illustrate the technique with a variety of applications, including tilings with tiles of irrational area, ordered integer factorizations, lattice walks enumerated by Euclidean length, and plane trees with vertices of irrational size. We also explore phase transitions in the asymptotics of families of irrational combinatorial classes., Comment: 27 pages

In Extreme Multi Label Completion (XMLCo), the objective is to predict the missing labels of a collection of documents. Together with XML Classification, XMLCo is arguably one of the most challenging document classification tasks, as the very high number of labels (at least ten of thousands) is generally very large compared to the number of available labelled documents in the training dataset. Such a task is often accompanied by a taxonomy that encodes the labels organic relationships, and many methods have been proposed to leverage this hierarchy to improve the results of XMLCo algorithms. In this paper, we propose a new approach to this problem, TAMLEC (Taxonomy-Aware Multi-task Learning for Extreme multi-label Completion). TAMLEC divides the problem into several Taxonomy-Aware Tasks, i.e. subsets of labels adapted to the hierarchical paths of the taxonomy, and trains on these tasks using a dynamic Parallel Feature sharing approach, where some parts of the model are shared between tasks while others are task-specific. Then, at inference time, TAMLEC uses the labels available in a document to infer the appropriate tasks and to predict missing labels. To achieve this result, TAMLEC uses a modified transformer architecture that predicts ordered sequences of labels on a Weak-Semilattice structure that is naturally induced by the tasks. This approach yields multiple advantages. First, our experiments on real-world datasets show that TAMLEC outperforms state-of-the-art methods for various XMLCo problems. Second, TAMLEC is by construction particularly suited for few-shots XML tasks, where new tasks or labels are introduced with only few examples, and extensive evaluations highlight its strong performance compared to existing methods.

The primary promise of decentralized learning is to allow users to engage in the training of machine learning models in a collaborative manner while keeping their data on their premises and without relying on any central entity. However, this paradigm necessitates the exchange of model parameters or gradients between peers. Such exchanges can be exploited to infer sensitive information about training data, which is achieved through privacy attacks (e.g Membership Inference Attacks -- MIA). In order to devise effective defense mechanisms, it is important to understand the factors that increase/reduce the vulnerability of a given decentralized learning architecture to MIA. In this study, we extensively explore the vulnerability to MIA of various decentralized learning architectures by varying the graph structure (e.g number of neighbors), the graph dynamics, and the aggregation strategy, across diverse datasets and data distributions. Our key finding, which to the best of our knowledge we are the first to report, is that the vulnerability to MIA is heavily correlated to (i) the local model mixing strategy performed by each node upon reception of models from neighboring nodes and (ii) the global mixing properties of the communication graph. We illustrate these results experimentally using four datasets and by theoretically analyzing the mixing properties of various decentralized architectures. Our paper draws a set of lessons learned for devising decentralized learning systems that reduce by design the vulnerability to MIA., Comment: 12 pages, 8 figures

The non-isothermal analysis of materials with the application of the Arrhenius equation involves temperature integration. If the frequency factor in the Arrhenius equation depends on temperature with a power-law relationship, the integral is known as the general temperature integral. This integral which has no analytical solution is estimated by the approximation functions with different accuracies. In this article, the rational approximations of the integral were obtained based on the minimization of the maximal deviation of bivariate functions. Mathematically, these problems belong to the class of quasiconvex optimization and can be solved using the bisection method. The approximations obtained in this study are more accurate than all approximates available in the literature., Comment: This paper was rejected from ARXIV, but we appealed and we were granted a permission to resubmit. "Thank you for your patience. Your appeal was accepted. You may resubmit this work to math.OC at your convenience."

To address the current crises (climatic, social, economic), the self-sufficiency -- a set of practices that combine energy sobriety, self-production of food and energy, and self-construction - arouses an increasing interest. The CNRS STAY project (Savoirs Techniques pour l'Auto-suffisance, sur YouTube) explores this topic by analyzing techniques shared on YouTube. We present Agro-STAY, a platform designed for the collection, processing, and visualization of data from YouTube videos and their comments. We use Natural Language Processing (NLP) techniques and language models, which enable a fine-grained analysis of alternative agricultural practice described online. -- Face aux crises actuelles (climatiques, sociales, \'economiques), l'auto-suffisance -- ensemble de pratiques combinant sobri\'et\'e \'energ\'etique, autoproduction alimentaire et \'energ\'etique et autoconstruction - suscite un int\'er\^et croissant. Le projet CNRS STAY (Savoirs Techniques pour l'Auto-suffisance, sur YouTube) s'inscrit dans ce domaine en analysant les savoirs techniques diffus\'es sur YouTube. Nous pr\'esentons Agro-STAY, une plateforme d\'edi\'ee \`a la collecte, au traitement et \`a la visualisation de donn\'ees issues de vid\'eos YouTube et de leurs commentaires. En mobilisant des techniques de traitement automatique des langues (TAL) et des mod\`eles de langues, ce travail permet une analyse fine des pratiques agricoles alternatives d\'ecrites en ligne., Comment: 8 pages, in French language, 3 figures

We present the results of the second commissioning phase of the short-focal-length area of the Apollon laser facility (located in Saclay, France), which was performed with the main laser beam (F1), scaled to a peak power of 2 PetaWatt. Under the conditions that were tested, this beam delivered on-target pulses of maximum energy up to 45 J and 22 fs duration. Several diagnostics were fielded to assess the performance of the facility. The on-target focal spot and its spatial stability, as well as the secondary sources produced when irradiating solid targets, have all been characterized, with the goal of helping users design future experiments. The laser-target interaction was characterized, as well as emissions of energetic ions, X-ray and neutrons recorded, all showing good laser-to-target coupling efficiency. Moreover, we demonstrated the simultaneous fielding of F1 with the auxiliary 0.5 PW F2 beam of Apollon, enabling dual beam operation. The present commissioning will be followed in 2025 by a further commissioning stage of F1 at the 8 PW level, en route to the final 10 PW goal.

We develop an extension of the Monte Carlo wave function approach that unambiguously identifies dynamical entanglement in general composite, open systems. Our algorithm performs tangential projections onto the set of separable states, leading to classically correlated quantum trajectories. By comparing this restricted evolution with the unrestricted one, we can characterize the entangling capabilities of quantum channels without making use of input-output relations. Moreover, applying this method is equivalent to solving the nonlinear master equation in Lindblad form introduced in \cite{PAH24} for two-qubit systems. We here extend these equations to multipartite systems of qudits, describing non-entangling dynamics in terms of a stochastic differential equation. We identify the impact of dynamical entanglement in open systems by applying our approach to several correlated decay processes. Therefore, our methodology provides a complete and ready-to-use framework to characterize dynamical quantum correlations caused by arbitrary open-system processes.

Providing entanglement for the design of quantum technologies in the presence of noise constitutes today's main challenge in quantum information science. A framework is required that assesses the build-up of entanglement in realistic settings. In this work, we put forth a new class of nonlinear quantum master equations in Lindblad form that unambiguously identify dynamical entanglement in open quantum systems via deviations from a separable evolution. This separability Lindblad equation restricts quantum trajectories to classically correlated states only. Unlike many conventional approaches, here the entangling capabilities of a process are not characterized by input-output relations, but separability is imposed at each instant of time. We solve these equations for crucial examples, thereby quantifying the dynamical impact of entanglement in non-equilibrium scenarios. Our results allow to benchmark the engineering of entangled states through dissipation. The separability Lindblad equation provides a unique path to characterizing quantum correlations caused by arbitrary system-bath interactions, specifically tailored for the noisy intermediate-scale quantum era.

We study the identifiability of nonlinear network systems with partial excitation and partial measurement when the network dynamics is linear on the edges and nonlinear on the nodes. We assume that the graph topology and the nonlinear functions at the node level are known, and we aim to identify the weight matrix of the graph. Our main result is to prove that fully-connected layered feed-forward networks are generically locally identifiable by exciting sources and measuring sinks in the class of analytic functions that cross the origin. This holds even when all other nodes remain unexcited and unmeasured and stands in sharp contrast to most findings on network identifiability requiring measurement and/or excitation of each node. The result applies in particular to feed-forward artificial neural networks with no offsets and generalizes previous literature by considering a broader class of functions and topologies., Comment: 9 pages, 2 figures

Plug-and-Play methods for image restoration are iterative algorithms that solve a variational problem to restore an image. These algorithms are known to be flexible to changes of degradation and to perform state-of-the-art restoration. Recently, a lot of efforts have been made to explore new algorithms to solve this variational problem based on the Plug-and-Play or REgularization by Denoising (RED) frameworks, such as SNORE that is a converging stochastic gradient descent algorithm. A variant of this algorithm, named SNORE Prox, reaches state-of-the-art performances, especially for inpainting tasks. However, the convergence of SNORE Prox, that can be seen as a stochastic proximal gradient descent, has not been analyzed so far. In this paper, we prove the convergence of SNORE Prox under reasonable non-convex assumptions.

The ``ratchet principle", which states that non-equilibrium systems violating parity symmetry generically exhibit steady-state currents, is one of the few generic results outside thermal equilibrium. We study exceptions to this principle observed in active and passive systems with spatially varying fluctuations sources. For dilute systems, we show that a hidden time-reversal symmetry prevents the emergence of ratchet currents. At higher densities, pairwise forces break this symmetry but an emergent conservation law for the momentum field may nevertheless prevent steady currents. We show how the presence of this conservation law can be tested analytically and characterize the onset of ratchet currents in its absence. Our results show that the ratchet principle should be amended to preclude parity symmetry, time-reversal symmetry, and bulk momentum conservation., Comment: 7 pages, 5 figures