March 2025 arXiv papers — page 155

Bhargav Acharya, William Saakyan, Barbara Hammer, Hanna Drimalla

Heart rate is a physiological signal that provides information about an individual's health and affective state. Remote photoplethysmography (rPPG) allows the estimation of this signal from video recordings of a person's face. Classical rPPG methods make use of signal processing techniques, while recent rPPG methods utilize deep learning networks. Methods ar

Minsu Kim, Rodrigo Mira, Honglie Chen, Stavros Petridis

In this paper, we investigate a novel approach for Target Speech Extraction (TSE), which relies solely on textual context to extract the target speech. We refer to this task as Contextual Speech Extraction (CSE). Unlike traditional TSE methods that rely on pre-recorded enrollment utterances, video of the target speaker's face, spatial information, or other e

CMS Collaboration

This paper presents the first search for a cH signal sensitive to the coupling of the charm quark (c) to the Higgs boson (H) in the associated production of at least one charm quark with a Higgs boson decaying to two photons. The results are based on a data set of proton-proton collisions at a center-of-mass energy of 13 TeV collected with the CMS experiment

Guanjie Huang, Danny Hin Kwok Tsang, Li Liu

Cued Speech (CS) is an innovative visual communication system that integrates lip-reading with hand coding, designed to enhance effective communication for individuals with hearing impairments. Automatic CS Recognition (ACSR) refers to the AI-driven process of automatically recognizing hand gestures and lip movements in CS, converting them into text. However

Seongho Son, William Bankes, Sangwoong Yoon, Shyam Sundhar Ramesh

We introduce Robust Multi-Objective Decoding (RMOD), a novel inference-time algorithm that robustly aligns Large Language Models (LLMs) to multiple human objectives (e.g., instruction-following, helpfulness, safety) by maximizing the worst-case rewards. RMOD formulates the robust decoding problem as a maximin two-player game between adversarially computed re

Yunke Ao, Johannes Köhler, Manish Prajapat, Yarden As

We propose a stochastic Model Predictive Control (MPC) framework that ensures closed-loop chance constraint satisfaction for linear systems with general sub-Gaussian process and measurement noise. By considering sub-Gaussian noise, we can provide guarantees for a large class of distributions, including time-varying distributions. Specifically, we first provi

Alejandro A. Hnilo

In the famous Solvay 1927 conference, Einstein discussed a gedanken experiment involving a single photon diffracted at an aperture and impinging on a screen. He devised the example to support De Broglie's hypothesis of the pilot wave, and his own ideas on the incompleteness of the description of physical reality provided by Quantum Mechanics. Partial realiza

Vinit Vijay Deshpande, Romana Piat

The work describes a numerical method to study the nature of compressive failure of porous ceramics in relation to its volume fraction. The microstructure of an alumina-based foam material manufactured by mechanical stirring of a slurry is studied here. A finite element-based compressive failure simulation of a real microstructure obtained from microtomograp

Julius F. Grimminger, Amihay Hanany, Deshuo Liu

This paper presents new developments in the study of 3d mirror symmetry and the phase structure of Abelian gauge theories. Previous works identified 3d mirrors for a specific class of theories, termed ``simple" Abelian theories. This work extends this framework by proposing 3d mirrors for ``non-simple" Abelian theories with both discrete and continuous gauge

Annastasia Haynie, Samantha C. Wu, Anthony L. Piro, Jim Fuller

Ultra-stripped and Type Ibn supernovae (USSNe and SNe Ibn, respectively) are fast-evolving, hydrogen-poor transients that often show signs of interaction with dense circumstellar material (CSM). Wu & Fuller (2022) identify a mass range for helium-core stars in which they expand significantly during core oxygen/neon burning, resulting in extreme late-stage ma

Aiden Sheckler

We resolve the existence of mixed 't Hooft anomalies between the electric and magnetic (solitonic) symmetries in $\sigma$-models and gauge theories. We identify the anomaly as naturally originating from a higher group in the Whitehead tower of the target space. In particular, the magnetic charges (topological solitons) can be interpreted as higher connection

Sun Woo P. Kim, Friedrich Hübner, Juan P. Garrahan, Benjamin Doyon

The emergence of hydrodynamics is one of the deepest phenomena in many-body systems. Arguably, the hydrodynamic equations are also the most important tools for predicting large-scale behaviour. Understanding how such equations emerge from microscopic deterministic dynamics is a century-old problem, despite recent progress in fine-tuned integrable systems. Du

Sagad Hamid, Tanya Braun

Efficient probabilistic inference by variable elimination in graphical models requires an optimal elimination order. However, finding an optimal order is a challenging combinatorial optimisation problem for models with a large number of random variables. Most recently, a reinforcement learning approach has been proposed to find efficient contraction orders i

Natalie B. Hogg, James W. Nightingale, Quihan He, Jacqueline McCleary

We compare forecasts for the abundance and properties of strong gravitational lenses in the COSMOS-Web survey, a $0.54$ deg$^2$ survey of the COSMOS field using the NIRCam and MIRI instruments aboard JWST, with the first catalogue of strong lens candidates identified in the observed NIRCam data, COWLS. We modify the lenspop package to produce a forecast for

Félix Desrochers, Mark R. Hirsbrunner, Joe Huxford, Adarsh S. Patri

Anomalous Hall crystals (AHCs) are exotic phases of matter that simultaneously break continuous translation symmetry and exhibit the quantum anomalous Hall effect. AHCs have recently been proposed to explain the observation of an integer quantum anomalous Hall phase in a multilayer graphene system. Despite intense theoretical and experimental interest, littl

Tom Vandekerckhove, Jasper De Witte, Lisa De Jaeger, Ewoud Vissers

The integration of second-order optical nonlinearities into scalable photonic platforms remains a key challenge due to their large sensitivity to fabrication variations. Here, we present a scalable quadratic nonlinear platform that harnesses the maturity and scalability of existing CMOS processes by heterogeneously integrating periodically poled lithium niob

Guillaume Mahler, James W. Nightingale, Natalie B. Hogg, Ghassem Gozaliasl

The COSMOS-Web Lens Survey (COWLS) presents the first systematic search for strong gravitational lenses in the COSMOS-Web field using data from the \textit{James Webb} Space Telescope (\textit{JWST}). Using high-resolution NIRCam imaging, we visually inspected over 42\,660 galaxies and identified over 400 lensing candidates. From this sample and based on \te

M. Cepeda, J. M. No, C. Ramos, R. M. Sandá Seoane

Exotic Higgs decays are among the most promising areas to be explored at the High-Luminosity LHC, given the unprecedentedly large amount $(\sim 3 \times 10^8)$ of 125 GeV Higgs bosons that will be produced. In this context, we propose a new search channel for which the Higgs boson decays to a (leptonically decaying) $Z$ boson and a light BSM pseudoscalar $a$

Biplob Bhattacherjee, Camellia Bose, Herbi K. Dreiner, Nivedita Ghosh

In the search for beyond the Standard Model (SM) physics, long-lived particles (LLPs) have emerged as potential candidates and are being explored in various ongoing experiments. Future lepton colliders, such as the FCC-ee, shall provide an excellent opportunity to probe LLPs, owing to their clean environment and improved particle identification. This study i

Fred Angelo Batan Garcia, Massimo Ricotti, Kazuyuki Sugimura

We model the formation of star clusters in a dwarf galaxy progenitor during the first $700 ~{\rm Myr}$ of cosmic history using a cosmological radiation-hydrodynamic simulation with a sub-grid star formation efficiency (SFE) model calibrated from AU-scale radiation-MHD simulations of molecular clouds with varying mass, density, and metallicity. In comparison

Christopher Cain, Anson D'Aloisio, Julian B. Munoz

Measurements of the ionisation state of the intergalactic medium (IGM) can probe the sources of the extragalactic ionising background. We provide new measurements of the ionising emissivity of galaxies using measurements of the ionising background and ionising photon mean free path from high-redshift quasar spectra at $2.5 < z < 6$. Unlike most prior works,

James Nightingale, Guillaume Mahler, Jacqueline McCleary, Qiuhan He

We present the COSMOS-Web Lens Survey (COWLS), a sample of over 100 strong lens candidates from the $0.54$\,deg$^2$ COSMOS-Web survey, discovered using exquisite James Webb Space Telescope (JWST) imaging across four wavebands. Following two rounds of visual inspection, over 100 candidates were ranked as `high confidence' or `likely' by at least $50\%$ of ins

Ruizhe Shen, Tianqi Chen, Bo Yang, Yin Zhong

Topology and symmetry play critical roles in characterizing quantum phases of matter. Recent advancements have unveiled symmetry-protected topological (SPT) phases in many-body systems as a unique class of short-range entangled states, notable for their nontrivial edge modes and characteristic ground-state entanglement gap. In this study, we demonstrate the

Tom Westerdijk, Chen Yang

We establish an exact correspondence between tree-level cosmological correlators with unparticle exchange (at integer scaling dimensions) and banana diagrams of conformally coupled scalars. This duality enables us to systematically solve the governing differential equations through the application of shift relations and de Sitter bootstrap techniques. Furthe

Dean Carmi, Javier Moreno, Shimon Sukholuski

Dispersion relations are nonperturbative formulas that relate the ultraviolet and infrared behavior of an observable with wide-ranging applications applications in linear response theory, quantum field theory scattering amplitudes, and conformal correlators. We derive a position-space dispersion relation for scalar four-point mixed correlation functions in a

Cesar A. Agón, Horacio Casini, Pedro J. Martinez

We explore the holographic prescription for computing the refined R\'enyi entropies $\tilde S_n$ in the $n \to 0$ limit within the AdS$_{d+1}$/CFT$_d$ framework. This limit can be interpreted as a high-temperature regime with respect to the energy defined by the modular Hamiltonian of the state reduced to a subregion. To leading order in $n$, we find that th

Jacob M. Leedom, Hitoshi Murayama, Gup Singh, Bethany Suter

We present exact results in softly-broken supersymmetric $\text{SU}(N_C)$ chiral gauge theories with charged fermions in one antisymmetric, $N_F$ fundamental, and $N_C+N_F-4$ anti-fundamental representations. We achieve this by considering the supersymmetric version of these theories and utilizing anomaly mediated supersymmetry breaking at a scale $m \ll \La

Franz Ciceri, Henning Samtleben

Holographic dualities that relate type II strings on near-horizon Dp brane geometries to super Yang-Mills theories with sixteen supercharges in p+1 dimensions provide non-conformal generalizations of the famous AdS/CFT correspondence. For the extremal case p=-1, this suggests a holographic duality for the IKKT matrix model - super Yang-Mills theory reduced t

Wenjun Niu, Victor Py

In this paper, we define (cohomologically) 1-shifted Manin triples and 1-shifted Lie bialgebras, and study their properties. We derive many results that are parallel to those found in ordinary Lie bialgebras, including the double construction and the existence of a 1-shifted $r$-matrix satisfying the classical Yang-Baxter equation. Turning to quantization, w

Ivan Medina, Sérgio R. Muniz, Elisa I. Goettems, Diogo O. Soares-Pinto

The optical pumping process is a fundamental tool for quantum protocols using the electronic and nuclear spins in the Nitrogen-Vacancy (NV) centers platform. In this work, we explore the process of electronic spin polarization through optical pumping from a thermodynamic perspective. We identify the sources of work and heat and show that the heat current is

S. J. Brennan, S. Barmentloo, S. Schulze, K. W. Smith

Over the past two decades, an increasing number of transients have shown luminous activity at their explosion sites weeks to years before an interacting supernova (SN) is observed. For some objects, this pre-SN activity is typically linked to large-scale mass-loss events preceding core collapse, yet its triggering mechanism and the underlying explosion proce

Joshuah T. Heath, Alexander C. Tyner, Thue Christian Thann, Vincent P. Michal

Superconducting qubits are often adversely affected by two-level systems (TLSs) within the Josephson junction, which contribute to decoherence and subsequently limit the performance of the qubit. By treating the TLS as a soft (i.e., low-frequency) bosonic mode localized in real space, we find that a single TLS in either the amorphous oxide surface or the sup

Yan G. Grange, Kevin Tai

In 2023, ASTRON took the step of incorporating a dedicated User Experience (UX) designer into its software development process. This decision aimed to enhance the accessibility and usability of services providing access to the data holdings from the telescopes we are developing. The field of astronomical software development has historically under emphasized

G. V. Pignataro, S. P. O'Sullivan, A. Bonafede, G. Bernardi

Magnetic fields in large scale structure filaments beyond galaxy clusters remain poorly understood. Superclusters offer a unique setting to study these low density environments, where weak signals make detection challenging. The Faraday rotation measure (RM) of polarized sources along supercluster lines of sight helps constrain magnetic field properties in t

Yongdong Luo, Wang Chen, Xiawu Zheng, Weizhong Huang

Recent advances in long video understanding typically mitigate visual redundancy through visual token pruning based on attention distribution. However, while existing methods employ post-hoc low-response token pruning in decoder layers, they overlook the input-level semantic correlation between visual tokens and instructions (query). In this paper, we propos

Ariba Khan, Stephen Casper, Dylan Hadfield-Menell

Research on the 'cultural alignment' of Large Language Models (LLMs) has emerged in response to growing interest in understanding representation across diverse stakeholders. Current approaches to evaluating cultural alignment through survey-based assessments that borrow from social science methodologies often overlook systematic robustness checks. Here, we i

Dohyeong Kim, Myungshin Im

In the merger-driven galaxy evolution scenario, the central supermassive black holes (SMBHs) in dust obscured galaxies grow rapidly. Interestingly, a recent work \citep{suh24} on a dust-obscured galaxy, LID-568 at $z=3.965$, has shown that its SMBH is growing extremely fast at about 40 times of the Eddington-limited accretion rate (i.e., super-Eddington accr

Jialv Zou, Bencheng Liao, Qian Zhang, Wenyu Liu

Recent advancements in unified multimodal understanding and visual generation (or multimodal generation) models have been hindered by their quadratic computational complexity and dependence on large-scale training data. We present OmniMamba, the first linear-architecture-based multimodal generation model that generates both text and images through a unified

Xin Wen, Bingchen Zhao, Ismail Elezi, Jiankang Deng

We introduce a novel visual tokenization framework that embeds a provable PCA-like structure into the latent token space. While existing visual tokenizers primarily optimize for reconstruction fidelity, they often neglect the structural properties of the latent space--a critical factor for both interpretability and downstream tasks. Our method generates a 1D

Rene A. Mendez, Andrei Tokovinin, Edgardor Costa, Max Dirk

We present measurements in the context of a survey of southern hemisphere binary and multiple stellar systems observed with the Zorro Speckle dual diffraction-limited optical imaging camera on the 8.1m Gemini-South telescope carried out between 2019 and 2023. The overall motivation of our survey, as well as some initial results of these observations, are out

Haoyu Wang, Sunhao Dai, Haiyuan Zhao, Liang Pang

Previous studies have found that PLM-based retrieval models exhibit a preference for LLM-generated content, assigning higher relevance scores to these documents even when their semantic quality is comparable to human-written ones. This phenomenon, known as source bias, threatens the sustainable development of the information access ecosystem. However, the un

Changxing Liu, Genjia Liu, Zijun Wang, Jinchang Yang

Vehicle-to-vehicle (V2V) cooperative autonomous driving holds great promise for improving safety by addressing the perception and prediction uncertainties inherent in single-agent systems. However, traditional cooperative methods are constrained by rigid collaboration protocols and limited generalization to unseen interactive scenarios. While LLM-based appro

Zhenfeng Ouyang, Rong-Qiang He, Zhong-Yi Lu

The discovery of superconductivity in Ruddlesden-Popper (RP) nickelates has drawn great attention. Many works have been done to study the superconductivity as well as to find more superconducting RP nickelates. However, there is a lack of general understanding regarding the key factors that contribute to the superconductivity of RP nickelates. Here, we syste

Viktor Moskvoretskii, Chris Biemann, Irina Nikishina

Although large language models (LLMs) have achieved remarkable performance across various tasks, they remain prone to errors. A key challenge is enabling them to self-correct. While prior research has relied on external tools or large proprietary models, this work explores self-correction in small language models (SLMs) through iterative fine-tuning using so

Nithin Parsan, David J. Yang, John J. Yang

Protein language models have revolutionized structure prediction, but their nonlinear nature obscures how sequence representations inform structure prediction. While sparse autoencoders (SAEs) offer a path to interpretability here by learning linear representations in high-dimensional space, their application has been limited to smaller protein language mode

Muzi Li, B. R. McNamara, Irina Zhuravleva

We measure surface brightness fluctuations in Chandra X-ray images of the cores of the galaxy clusters Abell 2029, Abell 2151, Abell 2107, RBS0533, and RBS0540. Their relatively structureless X-ray atmospheres exhibit the thermodynamic properties of cool cores including short central cooling times and low entropy. However, unlike typical cool-core clusters,

Yuanhao Wang, Cheng Zhang, Gonçalo Frazão, Jinlong Yang

We introduce GarmentCrafter, a new approach that enables non-professional users to create and modify 3D garments from a single-view image. While recent advances in image generation have facilitated 2D garment design, creating and editing 3D garments remains challenging for non-professional users. Existing methods for single-view 3D reconstruction often rely

Yongsheng Yu, Ziyun Zeng, Haitian Zheng, Jiebo Luo

Diffusion-based generative models have revolutionized object-oriented image editing, yet their deployment in realistic object removal and insertion remains hampered by challenges such as the intricate interplay of physical effects and insufficient paired training data. In this work, we introduce OmniPaint, a unified framework that re-conceptualizes object re

Jinchang Zhang, Zijun Li, Guoyu Lu

Depth-guided multimodal fusion combines depth information from visible and infrared images, significantly enhancing the performance of 3D reconstruction and robotics applications. Existing thermal-visible image fusion mainly focuses on detection tasks, ignoring other critical information such as depth. By addressing the limitations of single modalities in lo

Markus Heydenreich, Bas Lodewijks

We consider an evolving random discrete tree model called Preferential Attachment with Vertex Death, as introduced by Deijfen. Initialised with an alive root labelled $1$, at each step $n\geq1$ either a new vertex with label $n+1$ is introduced that attaches to an existing alive vertex selected preferentially according to a function $b$, or an alive vertex i

Tobias Kreiman, Aditi S. Krishnapriyan

Machine Learning Force Fields (MLFFs) are a promising alternative to expensive ab initio quantum mechanical molecular simulations. Given the diversity of chemical spaces that are of interest and the cost of generating new data, it is important to understand how MLFFs generalize beyond their training distributions. In order to characterize and better understa

Jiakai Lin, Jinchang Zhang, Guoyu Lu

Keypoint detection and local feature description are fundamental tasks in robotic perception, critical for applications such as SLAM, robot localization, feature matching, pose estimation, and 3D mapping. While existing methods predominantly operate on RGB images, we propose a novel network that directly processes raw images, bypassing the need for the Image

Christian Peterson, Edgar Gasperín, Alex Vañó-Viñuales

Numerical evolutions of a system of equations close to null infinity in the geometric background of the inversion-Minkowski spacetime are performed. The evolved equations correspond to the linearisation of a second order metric formulation of the conformal Einstein field equations (CEFEs). These linear equations were first presented in [1] and the main purpo

Amirhossein Samandar, Javier Carrón Duque, Craig J. Copi, Mikel Martin Barandiaran

We study the eigenmodes of the spin-2 Laplacian in orientable Euclidean manifolds and their implications for the tensor-induced part of the cosmic microwave background (CMB) temperature and polarization anisotropies. We provide analytic expressions for the correlation matrices of Fourier-mode amplitudes and of spherical harmonic coefficients. We demonstrate

Jasmine Li, Quanzhi Ye, Denis Vida, David L. Clark

The Taurid Complex is a large interplanetary system that contains comet 2P/Encke, several meteoroid streams, and possibly a number of near-Earth asteroids. The size and nature of the system has led to the speculation that it was formed through a large-scale cometary breakup. Numerical investigations have suggested that planetary dynamics can create a resonan

Zekun Li, Shinda Huang, Jiangtian Wang, Nathan Zhang

As language agents increasingly automate critical tasks, their ability to follow domain-specific standard operating procedures (SOPs), policies, and constraints when taking actions and making tool calls becomes essential yet remains underexplored. To address this gap, we develop an automated evaluation pipeline SOPBench with: (1) executable environments cont

Shuaiting Li, Juncan Deng, Chenxuan Wang, Kedong Xu

Vector Quantization (VQ) has emerged as a prominent weight compression technique, showcasing substantially lower quantization errors than uniform quantization across diverse models, particularly in extreme compression scenarios. However, its efficacy during fine-tuning is limited by the constraint of the compression format, where weight vectors assigned to t

Alexander V. Plavin, Yuri Y. Kovalev, Sergey V. Troitsky

Bright blazars were found to be prominent neutrino sources, and a number of IceCube events were associated with them over recent years. A particularly strong observational connection is present between neutrinos and blazars with bright, Doppler-boosted, parsec-scale radio emission. In this work, we further explore the nature of this connection by examining t

Xinxin Jiang

This paper provides evidence that stock returns, after truncation, might be modeled by a special type of continuous mixtures or normals, so-called $q$-Gaussians. Negative binomial distributions might model the counts for extreme returns. A generalized jump-diffusion model is proposed, and an explicit option pricing formula is obtained.

Yitian Zhang, Long Mai, Aniruddha Mahapatra, David Bourgin

We present a novel perspective on learning video embedders for generative modeling: rather than requiring an exact reproduction of an input video, an effective embedder should focus on synthesizing visually plausible reconstructions. This relaxed criterion enables substantial improvements in compression ratios without compromising the quality of downstream g

Yuhan Wang, Fangzhou Hong, Shuai Yang, Liming Jiang

Multiview diffusion models have shown considerable success in image-to-3D generation for general objects. However, when applied to human data, existing methods have yet to deliver promising results, largely due to the challenges of scaling multiview attention to higher resolutions. In this paper, we explore human multiview diffusion models at the megapixel l

Pierre Sermanet, Anirudha Majumdar, Alex Irpan, Dmitry Kalashnikov

Until recently, robotics safety research was predominantly about collision avoidance and hazard reduction in the immediate vicinity of a robot. Since the advent of large vision and language models (VLMs), robots are now also capable of higher-level semantic scene understanding and natural language interactions with humans. Despite their known vulnerabilities

Pierpaolo Basile, Lucia Siciliani, Elio Musacchio, Giovanni Semeraro

Word Sense Disambiguation (WSD) is a historical task in computational linguistics that has received much attention over the years. However, with the advent of Large Language Models (LLMs), interest in this task (in its classical definition) has decreased. In this study, we evaluate the performance of various LLMs on the WSD task. We extend a previous benchma

Yufeng Diao, Yichi Zhang, Daniele De Martini, Philip Guodong Zhao

This paper proposes a task-oriented co-design framework that integrates communication, computing, and control to address the key challenges of bandwidth limitations, noise interference, and latency in mission-critical industrial Cyber-Physical Systems (CPS). To improve communication efficiency and robustness, we design a task-oriented Joint Source-Channel Co

F. Magni, D. Riccobelli

The Plateau-Rayleigh instability shows that a cylindrical fluid flow can be destabilized by surface tension. Similarly, capillary forces can make an elastic cylinder unstable when the elastocapillary length is comparable to the cylinder's radius. While existing models predict a single isolated bulge as the result of an instability, experiments reveal a perio

Aliaksei Kachanovich, Jean Kimus, Steven Lowette, Michel H. G. Tytgat

Motivated by a small but intriguing excess observed in the decay mode $ H\rightarrow \ell^+\ell^- \gamma$ reported by both the ATLAS and CMS collaborations, we explore the possibility that new physics contributes directly to the effective $ H \ell \overline{\ell} \gamma $ coupling rather than modifying the $ Z $ peak. Concretely, we consider a dimension-8 op

A. N. Ormondroyd, W. J. Handley, M. P. Hobson, A. N. Lasenby

Recent cosmological surveys have provided unprecedented datasets that can be used to reconstruct the history of the dark energy equation of state. In this work, a free-form "flexknot'' parameterisation is employed to represent $w(a)$ as a linear spline between free-moving nodes, the number of which may vary. By combining DESI Baryon Acoustic Oscillation meas

Jackson R. Fliss

We give a prescription for minimally coupling massive matter to JT gravity with either sign of cosmological constant directly in its formulation as a topological BF theory. This coupling takes the form of a `Wilson spool,' originally introduced in the context of three-dimensional gravity. The Wilson spool expresses the exact one-loop partition function as th

Serena Federico, Davide Tramontana

In this work we study the smoothing effect of some variable coefficient operators of the form $D_t-A$, where $A$ is a Weyl-quantized pseudo-differential operator of order $m=2,3$. The class under consideration includes, among others, KdV-type and ultrahyperbolic Schr\"odinger operators. We prove homogeneous and inhomogeneous smoothing estimates and use them

Zihan Wang, Xinghao Qiao, Dong Li, Howell Tong

In this article, we propose a novel logistic quasi-maximum likelihood estimation (LQMLE) for general parametric time series models. Compared to the classical Gaussian QMLE and existing robust estimations, it enjoys many distinctive advantages, such as robustness in respect of distributional misspecification and heavy-tailedness of the innovation, more resili

M. Seetharama Gowda, David Sossa

We introduce the concepts of commutativity relative to a transformation group and strong commutativity in the setting of a semi-FTvN system and show their appearance as optimality conditions in certain optimization problems. In the setting of a semi-FTvN system (in particular, in an FTvN system), we show that strong commutativity implies commutativity and ob

Elliot S. Shannon, Andrew O. Finley, Paul B. May, Grant M. Domke

National forest inventory (NFI) data are often costly to collect, which inhibits efforts to estimate parameters of interest for small spatial, temporal, or biophysical domains. Traditionally, design-based estimators are used to estimate status of forest parameters of interest, but are unreliable for small areas where data are sparse. Additionally, design-bas

Wei Chen, Qiang Qiu

Data heterogeneity is one of the major challenges in federated learning (FL), which results in substantial client variance and slow convergence. In this study, we propose a novel solution: decomposing a convolutional filter in FL into a linear combination of filter subspace elements, i.e., filter atoms. This simple technique transforms global filter aggregat

Hu-Hua He, Mao-Jun Yan, Chun-Sheng An, Cheng-Rong Deng

We systematically explore the S-wave $D_1D_1$, $D_1D^*_2$ and $D^*_2D^*_2$ states with various isospin-spin-orbit ($ISL$) configurations in the quark model. We propose nine stable dimeson states with the $ISL$ configurations, $ISL=001$, $010$, $012$, $100$, $102$, $110$, $112$, $120$, and $122$, against dissociation into their constituent mesons. Those bound

Zhenchen Wan, Yanwu xu, Dongting Hu, Weilun Cheng

Recent advancements in Virtual Try-On (VITON) have significantly improved image realism and garment detail preservation, driven by powerful text-to-image (T2I) diffusion models. However, existing methods often rely on user-provided masks, introducing complexity and performance degradation due to imperfect inputs, as shown in Fig.1(a). To address this, we pro

Marta Calanchi, Massimo Grossi

In this paper we study the Poisson problem, \[ \begin{cases} -{\rm div}(d^\beta\nabla u)=f&{\rm in}\ \Omega\\ u=0&{\rm on}\ \partial\Omega, \end{cases} \] where $\Omega\subset\mathbb R^N$, $N\ge2$ is a smooth bounded domain, $f$ is a continuous function, $\beta< 1$, and $d(x)=dist(x,\partial\Omega )$. We describe the behaviour of $u$ near $\partial\Omega$ an

H. R Navod Thisura Peiris

Code suggestions have become an integral part of IDEs and developers use code suggestions generated by IDEs all the time. These code suggestions are mostly for calling a method of an object or for using a function of a library and not for possible next line of the code. GPT based models are too slow or resource intensive for real-time code suggestions in loc

Megan Masterson, Kishalay De, Christos Panagiotou, Erin Kara

Mid-infrared (MIR) emission from tidal disruption events (TDEs) is a powerful probe of the circumnuclear environment around dormant supermassive black holes. This emission arises from the reprocessing of intrinsic emission into thermal MIR emission by circumnuclear dust. While the majority of optical- and X-ray-selected TDEs show only weak dust echoes consis

Dmitri Bykov, Andrew Kuzovchikov

We consider isotropic and Lagrangian embeddings of coadjoint orbits of compact Lie groups into products of coadjoint orbits. After reviewing the known facts in the case of $\mathrm{SU}(n)$ we initiate a similar study for $\mathrm{SO}$ and $\mathrm{Sp}$ cases. In the second part we apply this to the study of dynamical systems with $\mathrm{SU}(n)$ symmetry, p

Anuj Aggarwal, Jorge Fernández-Pendás, Tahereh Abad, Daryoush Shiri

Flux-tunable qubits and couplers are common components in superconducting quantum processors. However, dynamically controlling these elements via current pulses poses challenges due to distortions and transients in the propagating signals. In particular, long-time transients can persist, adversely affecting subsequent qubit control operations. We model the f

Parishad BehnamGhader, Nicholas Meade, Siva Reddy

Instruction-following retrievers have been widely adopted alongside LLMs in real-world applications, but little work has investigated the safety risks surrounding their increasing search capabilities. We empirically study the ability of retrievers to satisfy malicious queries, both when used directly and when used in a retrieval augmented generation-based se

Zhenxin Zheng, Zhenjie Zheng

Curse of Dimensionality is an unavoidable challenge in statistical probability models, yet diffusion models seem to overcome this limitation, achieving impressive results in high-dimensional data generation. Diffusion models assume that they can learn the statistical quantities of the underlying probability distribution, enabling sampling from this distribut

Zecheng Zhang, Hao Liu, Wenjing Liao, Guang Lin

We propose a Coefficient-to-Basis Network (C2BNet), a novel framework for solving inverse problems within the operator learning paradigm. C2BNet efficiently adapts to different discretizations through fine-tuning, using a pre-trained model to significantly reduce computational cost while maintaining high accuracy. Unlike traditional approaches that require r

Sebastian Werner, Maria C. Borges, Karl Wolf, Stefan Tai

Current approaches to designing energy-efficient applications typically rely on measuring individual components using readily available local metrics, like CPU utilization. However, these metrics fall short when applied to cloud-native applications, which operate within the multi-tenant, shared environments of distributed cloud providers. Assessing and optim

Emily Xiao, Chin-Jou Li, Yilin Zhang, Graham Neubig

Many-shot in-context learning has recently shown promise as an alternative to finetuning, with the major advantage that the same model can be served for multiple tasks. However, this shifts the computational burden from training-time to inference-time, making deployment of many-shot ICL challenging to justify in-practice. This cost is further increased if a

Dušan Malić, Christian Fruhwirth-Reisinger, Samuel Schulter, Horst Possegger

LiDAR-based 3D detectors need large datasets for training, yet they struggle to generalize to novel domains. Domain Generalization (DG) aims to mitigate this by training detectors that are invariant to such domain shifts. Current DG approaches exclusively rely on global geometric features (point cloud Cartesian coordinates) as input features. Over-reliance o

Ruibin Yuan, Hanfeng Lin, Shuyue Guo, Ge Zhang

We tackle the task of long-form music generation--particularly the challenging \textbf{lyrics-to-song} problem--by introducing YuE, a family of open foundation models based on the LLaMA2 architecture. Specifically, YuE scales to trillions of tokens and generates up to five minutes of music while maintaining lyrical alignment, coherent musical structure, and

Adnan Ali Ahmad, Matthias González, Patrick Hennebelle, Ugo Lebreuilly

Although protostars and disks are often studied separately owing to numerical and observational challenges, breakthroughs in recent years have highlighted the need to study both objects in concert. The role of magnetic fields in this regard must be investigated. We aim to describe the birth of the protostar and that of its disk, as well as their early joint

Hubert Baniecki, Przemyslaw Biecek

A common belief is that intrinsically interpretable deep learning models ensure a correct, intuitive understanding of their behavior and offer greater robustness against accidental errors or intentional manipulation. However, these beliefs have not been comprehensively verified, and growing evidence casts doubt on them. In this paper, we highlight the risks

Christopher E. Latchford, Daniel L. Maser

We designed and built an adjustable repetition rate optical frequency comb using off-the-shelf components with several improvements to an established erbium fiber design. This design, built fully by an undergraduate, was assembled at a fraction of the cost of a commercial comb. This comb has a repetition rate of roughly 100 MHz with a tuning range of 3 MHz.

Vamoua Yachongka, Rémi A. Chou

This study investigates secret-key generation for device authentication using physical identifiers, such as responses from physical unclonable functions (PUFs). The system includes two legitimate terminals (encoder and decoder) and an eavesdropper (Eve), each with access to different measurements of the identifier. From the device identifier, the encoder gen

Wolfdieter Lang

In Descartes' five circle problem integer curvatures (inverse radii) are considered. The positive integer curvature triple [c_1, c_2, c_3] (dimensionless), with non-decreasing entries for three given mutually touching circles, leading to integer curvatures [c_{4,-}, c_{4,+}] for the two circles touching the given ones is called a Descartes-Steiner triple. Th

Adlin Abell-Ball, Elizabeth Gillaspy, George Glidden-Handgis, S. Joseph Lippert

We use the lens of Zappa--Sz\'ep decomposition to examine the relationship between directed graph products and $k$-graph products. There are many examples of higher-rank graphs, or $k$-graphs, whose underlying directed graph may be factored as a product, but the $k$-graph itself is not a product. In such examples, we establish that the Zappa--Sz\'ep structur

Zak Buzzard

Diffusion models have seen immense success in modelling continuous data across a range of domains such as vision and audio. Despite the challenges of adapting diffusion models to discrete data, recent work explores their application to text generation by working in the continuous embedding space. However, these models lack a natural means to control the inhe

Joseph Jachinowski, Hassan Alnatah, David W. Snoke, Peter B. Littlewood

Due to their photonic components, exciton-polariton systems provide a convenient platform to study the coherence properties of weakly-interacting Bose gases. In particular, optical interferometry enables the measurement of the first-order coherence function which provides information about the intrinsic correlations of the system. In this paper, we derive a

Satrio Adi Rukmono, Filip Zamfirov, Lina Ochoa, Floris Pex

Software architecture knowledge transfer is essential for software development, but related documentation is often incomplete or ambiguous, making oral explanations a common means. Our broader aim is to explore how such explanations might be better supported and eventually automated; as a prerequisite, we first investigate how explanations are actually condu

Aida Abiad, Nils Van de Berg, Robin Simoens

Switching is an operation on a graph that does not change the spectrum of the adjacency matrix, thus producing cospectral graphs. An important activity in the field of spectral graph theory is the characterization of graphs by their spectrum. Hence, switching provides a tool for disproving the existence of such a characterization. This paper presents a gener

Aleksandr Berezutskii, Minzhao Liu, Atithi Acharya, Roman Ellerbrock

In the rapidly evolving field of quantum computing, tensor networks serve as an important tool due to their multifaceted utility. In this paper, we review the diverse applications of tensor networks and show that they are an important instrument for quantum computing. Specifically, we summarize the application of tensor networks in various domains of quantum

Muzhi Zhu, Yuzhuo Tian, Hao Chen, Chunluan Zhou

While MLLMs have demonstrated adequate image understanding capabilities, they still struggle with pixel-level comprehension, limiting their practical applications. Current evaluation tasks like VQA and visual grounding remain too coarse to assess fine-grained pixel comprehension accurately. Though segmentation is foundational for pixel-level understanding, e

Mariola Klusek-Gawenda, Victor P. Goncalves, Antoni Szczurek

The current state-of-the-art theoretical estimations lead to cross-sections for $AA \to \gamma \gamma AA$ which are somewhat smaller than the measured ones by the ATLAS and CMS Collaborations, which motivates the searching and calculation of subleading corrections disregarded in these previous studies. In this paper, we estimate for the first time the contri