March 2026 arXiv papers — page 71

Sergio Cruz-Blázquez, Azahara DelaTorre

We derive Kazdan-Warner type identities for the boundary problem of prescribing nonconstant interior $Q$ curvature and boundary $T$ curvature on the upper hemisphere $\mathbb{S}^4_+$ by a conformal change of the standard metric. Using the natural variational formulation and conformal variations generated by boundary-preserving conformal vector fields, we obt

Michele Boreale, Luisa Collodi

We study provably correct and efficient instantiations of Sequential Monte Carlo (SMC) inference in the context of formal operational semantics of Probabilistic Programs (PPs). We focus on universal PPs featuring sampling from arbitrary measures and conditioning/reweighting in unbounded loops. We first equip Probabilistic Program Graphs (PPGs), an automata-t

T. Yu. Magakian, T. A. Movsessian, A. V. Moiseev, T. S. Molyarova

Optical spectra of the well-known infrared source CPM~19, which exhibited a strong decline in brightness during the period from 1984--1987 to 2000--2005, have been obtained for the first time. A strong and broad H$\alpha$ emission line has been detected, along with the possible presence of [S II] emission. No traces of an absorption spectrum are observed. It

Gonzalo Morras, Geraint Pratten, Patricia Schmidt

We revisit the population properties of neutron star-black hole (NSBH) mergers using low-mass compact binary coalescences reported through GWTC-4. Employing pyEFPE, an inspiral-only waveform model that captures both orbital eccentricity and spin-induced precession, we reanalyse all binary neutron star (BNS) and NSBH events observed via gravitational waves. T

Chi Wang, David Angeli

This paper develops a method to construct robust positively invariant (RPI) tube sets from finite noisy input-state data of an unknown linear time-invariant (LTI) system, yielding tubes that can be directly embedded in tube-based robust data-driven predictive control. Data-consistency uncertainty sets are constructed under process/measurement noise with poly

Yehudit Aperstein, Linoy Halifa, Sagiv Bar, Alexander Apartsin

Enhancing reader engagement while preserving informational fidelity is a central challenge in controllable text generation for news media. Optimizing news headlines for reader engagement is often conflated with clickbait, resulting in exaggerated or misleading phrasing that undermines editorial trust. We frame clickbait not as a separate stylistic category,

Hejun Dong, Junbo Niu, Bin Wang, Weijun Zeng

Optical character recognition (OCR) has evolved from line-level transcription to structured document parsing, requiring models to recover long-form sequences containing layout, tables, and formulas. Despite recent advances in vision-language models, most existing systems rely on autoregressive decoding, which introduces sequential latency and amplifies error

Aurélien Faucher

We prove the weak relative Kawamata-Morrison movable cone conjecture for K-trivial fibrations whose very general fibre is a quotient, by a finite group of automorphisms acting freely in codimension one, of a product of certain Calabi-Yau pairs whose underlying varieties have well-clipped movable cones, a notion recently introduced by C\'ecile Gachet. Our mai

Lorenzo Battini, Marko Milenković

We describe the algorithms used by the ETH Flippers team in the CG:SHOP 2026 Challenge. Each instance consists of a set of triangulations on a common point set, and the objective is to find a central triangulation that minimizes the total parallel flip distance to the input set. Our strategy combines an exact solver for small and medium-sized instances with

Valentina Genesini, Giacomo Galloni, Luca Pagano, Paolo Campeti

The Thomson scattering optical depth to reionisation, $\tau$, one of the six parameters of the $\Lambda$CDM model, is primarily constrained by the large-scale E-mode polarisation of the Cosmic Microwave Background (CMB). In this work, we present the E-mode Likelihood for Cross-Analysis (elica), a multi-frequency, harmonic-space likelihood that combines all c

Carlo Marzo

The consistent construction of quantum field theories beyond the simplest cases requires a precise characterization of the propagating degrees of freedom. These are encoded in the single-pole structure of two-point functions, a connection firmly established through foundational theoretical work of the last century. While high-level programs for spectral anal

Qifei Cui, Patrick Chen

Egocentric videos present unique challenges for 3D reconstruction due to rapid camera motion and frequent dynamic interactions. State-of-the-art static reconstruction systems, such as MapAnything, often degrade in these settings, suffering from catastrophic trajectory drift and "ghost" geometry caused by moving hands. We bridge this gap by proposing a robust

Javier Carrón Duque, Mikel Martin Barandiaran, Joseba Martínez-Arrizabalaga

The Cosmological Principle assumes a statistically isotropic Universe, but the Cosmic Microwave Background (CMB) exhibits some anomalous statistical features, such as the hemispherical power asymmetry, that challenge this core assumption. We aim to expand the characterization of this anomaly by investigating the uniformity of the CMB's morphological and topo

Zhiyao Wang, Aodhán Corrigan, Norbert Lütkenhaus

In phase-randomized weak coherent pulse (WCP) implementations of Quantum Key Distribution (QKD) BB84 protocol, the decoy method is often used to compensate BB84's vulnerability against photon number splitting (PNS) attacks. However, this typically introduces extra complexities and requirements on experimental devices. In this paper, we are therefore interest

Jiaying Zhu, Lyuye Zhang, Wenbo Guo, Yang Liu

An agent skill is a reusable package for extending an LLM agent, typically a SKILL.md file that combines YAML metadata, natural-language instructions, and executable code. Public repositories now host over two million skills, yet existing tools analyze each artifact in isolation, and registries do not track reuse created through copying, renaming, or adaptat

Haoming Meng, Kexin Huang, Shaohang Wei, Chiyu Ma

Reinforcement learning with verifiable rewards (RLVR) has significantly improved reasoning in large language models (LLMs), yet the token-level mechanisms underlying these improvements remain unclear. We present a systematic empirical study of RLVR's distributional effects organized around three main analyses: (1) token-level characterization of distribution

Anni Li, Yingqing Chen, Christos G. Cassandras, Wei Xiao

This paper studies the problem of finite-time convergence to a prescribed safe set for nonlinear systems whose initial states violate the safety constraints. Existing Control Lyapunov-Barrier Functions (CLBFs) can enforce recovery to the safe set but may suffer from the issue of chattering and they do not explicitly consider control bounds. To address these

GRAVITY Collaboration, K. Perraut, J. Bouvier, H. Nowacki

Protoplanetary disks around young Sun-like stars are the cradles of the vast majority of detected exoplanets. Probing these disks at multiple spatial scales is key to uncovering how planets form. We aim to spatially and spectrally resolve the inner disk and star-disk interaction region of the M0.3 T Tauri star DO Tau by combining two complementary techniques

Dubois David, Guichard Pierre, Pasquier Remi

Ubiquitously found in the Universe, atomic hydrogen represents up to 70% of the neutral gas composition of the Milky Way. As an adatom, hydrogen can physisorb or chemisorb onto interstellar dust grains and icy mantles, thereby contributing to the formation of H2 and, potentially, to the synthesis of more complex hydrogenated species. In addition, structures

Pragnya Das

We investigate the metric structure of the intersection lattice L(B(n,k)) of the discriminantal arrange ment using circuit supports. We show that the cover graph associated with L(B(n,k)) is isometrically embedded into a hypercube, making it a partial cube and a median graph, with distances given by the Hamming distance and geodesics described by symmetric d

Roberto Lionello, Cooper Downs, Emily I. Mason, Jon A. Linker

Magnetic flux fills the heliosphere, expands outward from the solar corona, and is fundamentally related to the structure and dynamics of the solar corona and solar wind. Open magnetic flux and the fast wind are thought to originate from open magnetic field lines in coronal holes. Less understood processes in the streamer belt and the boundaries of coronal h

Adam Nahum, Sthitadhi Roy

Deterministic classical cellular automata can be in two phases, depending on how irreversible the dynamical rules are. In the strongly irreversible phase, trajectories with different initial conditions coalesce quickly, while in the weakly irreversible phase, trajectories with different initial conditions can remain different for a time exponential in the sy

Marc Kegel, Jonathan Spreer

Two distinct knots are said to be friends if their complements, filled along the 0-slope, produce diffeomorphic 3-manifolds. In this article, we develop a practical algorithm, implemented using SnapPy and Regina, to search for a friend of a given knot. As an application, we construct a census of simple knots that admit friends and use these data to formulate

Tanvir Ahmed, Yixuan Gao, Adnan Armouti, Rajalakshmi Nandakumar

We present mmFHE, the first system that enables fully homomorphic encryption (FHE) for end-to-end mmWave radar sensing. mmFHE encrypts raw range profiles on a lightweight edge device and executes the entire mmWave signal-processing and ML inference pipeline homomorphically on an untrusted cloud that operates exclusively on ciphertexts. At the core of mmFHE i

Frederick Reiber

In this short position paper, I develop a dialectical framework for understanding the political ideology of technological projects. To do so, I draw on critical and emancipatory social science discussions, highlighting how both a project's values and constraints are necessary for understanding its ideology. A brief example is then presented to aid comprehens

Rohan Jha, Chunsheng Zuo, Reno Kriz, Benjamin Van Durme

While multi-vector retrieval models outperform single-vector models of comparable size in retrieval quality, their practicality is limited by substantially larger index sizes, driven by the additional sequence-length dimension in their document embeddings. Because document embedding size dictates both memory overhead and query latency, compression is essenti

Sam Bennett, Amihay Hanany, Rudolph Kalveks

This paper explores 3d $\mathcal{N}=4$ quiver gauge theories whose moduli spaces represent nilpotent orbits, S\l odowy slices or, more generally, S\l odowy intersections, which span the Special Pieces of nilcones of Classical or Exceptional algebras. We introduce a map between magnetic and electric quivers containing symmetric group actions, such as wreathin

Fateme Memar, Tao Zhe, Dongjie Wang

Symbolic regression aims to discover human-interpretable equations that explain observational data. However, existing approaches rely heavily on discrete structure search (e.g., genetic programming), which often leads to high computational cost, unstable performance, and limited scalability to large equation spaces. To address these challenges, we propose SR

Indranil Bhattacharyya, Rakesh Halder, Nir Lazarovich, Mahan Mj

Let $Y\to X$ be a proper map between proper hyperbolic metric spaces. A Cannon--Thurston map is a continuous extension $\partial Y \to \partial X$. We prove that in most known settings in which a Cannon--Thurston map exists it is uniformly finite-to-one. This answers a question due to Swarup from Bestvina's problem list and generalizes previous results of Ca

Shubhayan Sarkar

In the current world, the use of artificial intelligence is penetrating every aspect of human life. The basic element of any artificial intelligence is a digital neuron, called a perceptron, while its quantum analogue is called a quantum perceptron. Here, we introduce a model of perceptron called the informationally-restricted measurement-based perceptron (I

Jiayin Dong, Eve J. Lee, Eiichiro Kokubo, Ruth Murray-Clay

Warm giant planets with orbital periods of tens of days exhibit a positive correlation between mass and eccentricity. We interpret this trend as the outcome of planet-planet scattering, representing a transition from collision-dominated interactions among low-mass planets to ejection-dominated interactions among high-mass planets. This framework has importan

Johannes Buchner, Isabelle Gauger, Qiaoya Wu, Hattie Starck

In most galaxies, the central black hole accounts for no more than a percent of the total mass in stars. Recently, however, extremely over-massive black holes with ratios of 10% have been reported in dwarf galaxies at z<1 and at cosmic dawn (z>5.5) by JWST. Both findings have been interpreted as signatures of the still mysterious origins of super-massive bla

Yariv Kafri, Julien Tailleur

We study the simplest terms that need to be included in active field theories to couple them to external potentials. To do so, we consider active Brownian particles and implement a systematic perturbative expansion in the particle persistence time. The result is a non-trivial coupling between density and potential gradients, which accounts for the nonequilib

Piotr Bargiela, Amlan Chakraborty, Giulio Gambuti, Melih A. Ozcelik

We present the calculation of three-loop massless QCD and QED helicity amplitudes for light-by-light scattering. We make use of Lorentz tensor decomposition in the 't Hooft-Veltman dimensional regularisation scheme to reduce the complexity of the computation. Our analytic amplitude results are remarkably compact and can be efficiently evaluated numerically.

Erik J. Gustafson, Henry Lamm

We present an extension of the Evolving density matrices on Qubits (E$\rho$OQ) framework that enables efficient fault-tolerant preparation of fermionic quantum states. The original method circumvents state preparation by stochastic sampling, but faces a sign problem in fermionic systems leading to a large number of circuits necessary. We resolve this by comb

Alperen Keles, Ethan Chou, Harrison Goldstein, Leonidas Lampropoulos

Database management systems (DBMSs) are notoriously complex, making them difficult to test effectively, especially during early development when many features are incomplete. Traditional testing tools like SQLancer and SQLSmith are highly effective for mature databases, but they struggle with high false positive rates and low actionability when applied to ev

Alex Salvatierra, José Antonio Sanz, Christian Gutiérrez, Mikel Galar

Semantic segmentation metrics for 3D point clouds, such as mean Intersection over Union (mIoU) and Overall Accuracy (OA), present two key limitations in the context of aerial LiDAR data. First, they treat all misclassifications equally regardless of their spatial context, overlooking cases where the geometric severity of errors directly impacts the quality o

Cecilia Romero Rodríguez, Pau Amaro Seoane

The coalescence of degenerate helium cores during red giant collisions - a process we term erythrohenosis - introduces a novel class of transient astrophysical sources of high-energy neutrinos. Using stellar models generated with MESA and SPH simulations of the final inspiral phase, we develop a semi-analytical model to estimate the amount of hydrogen mixed

John Ellis, Jason L. Evans, Shihwen Hor, Natsumi Nagata

We examine the impact on the phenomenology of the minimal supersymmetric SU(5) Grand Unified Theory (GUT) of dimension-5 operators with coefficients suppressed by the Planck mass scale, with particular emphasis on predictions for nucleon decay. We incorporate dimension-5 operators in both the Higgs sector and the Yukawa interactions in the theory, and take a

Jianghui Xu, Jiang-Tao Li, Guilin Liu, Luan Luan

We present the VLA C-array S-band (2--4 GHz) radio continuum observations of the nearby edge-on spiral galaxy NGC 4565, a target from the Continuum Halos in Nearby Galaxies - an EVLA (CHANG-ES) Survey. We conduct rotation measure synthesis to probe the magnetic field structure and analyze the vertical radio continuum intensity profiles using the 1-D cosmic r

Vaibhav Sharma, Shung-An Koh, Jonathan Stepp, Dasom Kim

We study magnetic materials whose low energy physics can be effectively described by a Dicke model, which we term Dicke materials. We show how a Dicke model emerges in such materials due to a coexistence of fast-dispersing and slow-dispersing spins, which are strongly coupled. Analogous to the paradigmatic Dicke model describing light-matter interactions, th

Eugene Chen, Daniel Green, Vincent S. H. Lee

Primordial non-Gaussianity is one of the most powerful probes of the inflationary epoch. The particle spectrum relevant to inflation, including masses and spins, is encoded in the precise form of statistical correlations of the adiabatic modes. Yet, in the presence of nonlinear structure formation, the optimal approach to measuring these signals remains uncl

Xiaolin Ma, Volodymyr Takhistov, Hideo Iizuka

Casimir force searches provide among the most sensitive laboratory probes of new short range interactions. Existing constraints rely almost exclusively on a single geometry. We show that Casimir geometry constitutes an independent observable, as Yukawa-type interactions and Casimir background exhibit different geometric scaling for bulk forces and surface qu

Nasser M. Ahmed, Remziye Canbay, Deniz Cennet Çınar

NGC 2168 (M35) serves as a fundamental benchmark for studying stellar evolution and dynamical environments at the transition between young and intermediate-age populations. We present a comprehensive analysis of the cluster's kinematic, structural, and astrophysical properties utilizing high-precision astrometry and photometry data from Gaia Data Release 3 (

Julia Esposito, Gongjie Li, Songhu Wang

Recent observations show distinct orbital architectures for hot and warm Jupiters: hot Jupiters span a wide range of stellar obliquities and tend to host distant companions without close-by companions, whereas warm Jupiters are often aligned and accompanied by both close-by and distant companions. In this paper, we revisit planet-planet scattering and demons

Ta-Hsin Li

Spline quantile regression (SQR) is a method introduced recently by Li and Megiddo (2026) for linear quantile regression where the regression coefficients are treated as smooth functions of the quantile level. With the coefficients represented by cubic splines with fixed knots on a given set of quantiles, the SQR method produces an estimate for the functiona

Giovanni De Crescenzo, Javier Mariño Villadamigo, Nina Elmer, Theo Heimel

We combine fast amplitude surrogates with neural importance sampling to accelerate NLO calculations. For virtual corrections, a learned ratio to the Born matrix element with calibrated uncertainties guarantees reliable precision across phase space. For real emission, we stick to the standard FKS subtraction and train sector-conditioned surrogates of the regu

David Shlivko, Vivian Poulin

Recent analyses combining data from the cosmic microwave background (CMB), baryon acoustic oscillations (BAO), and Type Ia supernovae (SN) have revealed a tentative observational preference for phantom crossing in the dark energy equation of state $w$. We argue that this preference is a natural consequence of the $\Omega_m$ tensions that arise when these dat

Fabio Fontanot, Gabriella De Lucia, Lizhi Xie, Stefano Zibetti

Recent results form the James Webb Space Telescope (JWST) report space densities for bright and massive galaxies at z>7 that far exceed expectations of theoretical models of galaxy formation, prompting a revision of our understanding of the physical processes leading to the assembly of the first luminous structures. In this work we present predictions from a

Ricardo Olmedo, Bernhard Schölkopf, Moritz Hardt

Consider a market of competing model providers selling query access to models with varying costs and capabilities. Customers submit problem instances and are willing to pay up to a budget for a verifiable solution. An arbitrageur efficiently allocates inference budget across providers to undercut the market, thus creating a competitive offering with no model

Georgios F. Paraschos, Ioannis Liodakis, Svetlana Jorstad, Yuri Y. Kovalev

The location of $\gamma$-ray creation and emission within extra-galactic jets is a matter of active debate. One particularly well-suited source to pinpoint the location is the nearby, bright radio galaxy 3C 84, harbouring a powerful jet. Here we investigate the origin of $\gamma$-rays measured during a recent $\gamma$-ray flare, by analysing the linear polar

Nicolaie Cernov, Jorge Íñiguez-González, Hugo Aramberri

Several candidate reference phases have been proposed to discuss phase transitions and ferroelectricity in hafnia in recent years. Although these proposals comply with crystallographic requirements, a physically compelling rationale connecting parent and daughter phases is often lacking. This problem is aggravated by the absence of clearly dominant polarizat

Vasilis Belis, Giulio Crognaletti, Matteo Argenton, Michele Grossi

Quantum computers provide a super-exponential speedup for performing a Fourier transform over the symmetric group, an ability for which practical use cases have remained elusive so far. In this work, we leverage this ability to unlock spectral methods for machine learning over permutation-structured data, which appear in applications such as multi-object tra

Nickolay Y. Gnedin, Harley Katz

The largest existing simulations of cosmic reionization model radiative transfer with moment methods that require a closure relation. The two most commonly used closure relations are M1 and OTVET; both close the moment hierarchy at the first moment. We explore the properties of a higher, second-order closure. We show that direct generalizations of M1 and OTV

Julien Baglio, Yacine Haddad, Richard Polifka

The development of new drugs is a tedious, time-consuming, and expensive process, for which the average costs are estimated to be up to around $2.5 billion. The first step in this long process is the design of the new drug, for which de novo drug design, assisted by artificial intelligence, has blossomed in recent years and revolutionized the field. In parti

Laurens Smulders, Johannes Noller

Dynamical dark energy theories generically introduce a time-dependent field that causes the accelerated expansion of the Universe on large scales. When embedding black hole solutions in such a cosmological spacetime, this time dependence naturally gives rise to cosmological hair, i.e. the local black hole physics is no longer controlled by just the mass and

Nabeen Bhusal, Simone Blasi, Thomas Konstandin, Enrico Perboni

We study slow-down effects for bubbles formed in a cosmological first-order phase transition (PT) focusing on deflagrations and hybrids, where the bubble wall is preceded by a shockwave of heated plasma. Slow-down has been observed in multi-bubble simulations together with a suppression of gravitational wave (GW) emission, mostly for slow walls. We study the

Yu-Min Hu, Yu-Bo Shi, Linhu Li, Gianluca Teza

Boundary perturbations are generally irrelevant for bulk properties in the thermodynamic limit, as they are edge-confined and subextensive. We show that this expectation breaks down in boundary-driven systems exhibiting the non-Hermitian skin effect, where arbitrarily weak boundary Floquet driving reconstructs bulk quasienergy spectra and dynamics. We develo

Kausik Ghosh, Miguel F. Paulos, Noé Suchel, Zechuan Zheng

We study sum rules that control the Regge limit of one-dimensional conformal field theory (CFT) correlators and relate them to dual bulk scattering processes at high energies in $\mathrm{AdS}_2$. By imposing the condition that no scattering takes place in the bulk, these sum rules single out special solutions to crossing symmetry that describe long-range mod

Kyu Jung Bae, Dhong Yeon Cheong, Jinn-Ouk Gong, Keisuke Harigaya

We formulate a new class of primordial perturbations called $\textit{kinetic isocurvature perturbations}$, where the mass density of dark matter is constant relative to the photon number density while the kinetic energy of dark matter fluctuates in space. Such perturbations naturally arise in scenarios where a nonrelativistic heavy field decays into relativi

Teodora-Elena Bulichi, Gene C. K. Leung, Anna-Christina Eilers, Pablo G. Perez-Gonzalez

Obscured active galactic nuclei (AGN) are often invoked to explain the rapid emergence of young quasars at high redshift and are crucial for building a complete census of AGN activity and black hole growth. The advent of the James Webb Space Telescope (JWST) extends the discovery space for obscured AGN into the mid-infrared (mid-IR) with unprecedented precis

Miguel Escudero, Maksym Ovchynnikov, Neal Weiner

The vast majority of extensions of the Standard Model affecting the number of effective relativistic neutrino species ($N_{\rm eff}$) do so additively, namely, they enhance this quantity with some light state contributing to dark radiation. In this work, we consider precisely the opposite case: new physics scenarios that can lead to $N_{\rm eff} < 3$ that ar

Itay M. Bloch, Simon Knapen, Xinran Li, Amalia Madden

We present a concept for a tabletop-scale detector with an amorphous target designed to search for dark matter absorption into phonon excitations. In crystalline materials, absorption occurs only at narrow resonances where the dark matter mass matches a zero momentum optical phonon mode, whereas amorphous targets provide a broadband response that can substan

Tevž Lotrič, Steven H. Simon

We study phases of itinerant anyons when hole-doping Laughlin-like states in fractional Chern insulators (FCIs). In light of the recent observation of time-reversal-broken superconductivity near FCIs in van der Waals materials, a theoretical understanding of doped fractional quantum Hall states on a lattice has been developed by Shi and Senthil [Phys. Rev. X

Hao Chen, Pier Francesco Monni, Zhaoyan Pang, Gherardo Vita

Understanding the link between correlation functions (CFs) of local operators and measurable collider correlators has emerged as a new opportunity in the study of gauge theory dynamics at colliders. While in Conformal Field Theories (CFTs) this connection is established by the light transform, the non-conformal nature of QCD complicates its use beyond the lo

Umair Nawaz, Ahmed Heakl, Ufaq Khan, Abdelrahman Shaker

Diffusion Transformers (DiTs) power high-fidelity video world models but remain computationally expensive due to sequential denoising and costly spatio-temporal attention. Training-free feature caching accelerates inference by reusing intermediate activations across denoising steps; however, existing methods largely rely on a Zero-Order Hold assumption i.e.,

Ruoliu Yang, Chu Wu, Caifeng Shan, Ran He

Long video understanding remains challenging for multimodal large language models (MLLMs) due to limited context windows, which necessitate identifying sparse query-relevant video segments. However, existing methods predominantly localize clues based solely on the query, overlooking the video's intrinsic structure and varying relevance across segments. To ad

Shivam Duggal, Xingjian Bai, Zongze Wu, Richard Zhang

Latent diffusion models (LDMs) enable high-fidelity synthesis by operating in learned latent spaces. However, training state-of-the-art LDMs requires complex staging: a tokenizer must be trained first, before the diffusion model can be trained in the frozen latent space. We propose UNITE - an autoencoder architecture for unified tokenization and latent diffu

Zhide Zhong, Junfeng Li, Junjie He, Haodong Yan

Vision-Language-Action (VLA) models map visual observations and language instructions directly to robotic actions. While effective for simple tasks, standard VLA models often struggle with complex, multi-step tasks requiring logical planning, as well as precise manipulations demanding fine-grained spatial perception. Recent efforts have incorporated Chain-of

Haoyu Zhen, Xiaolong Li, Yilin Zhao, Han Zhang

Large Language Models (LLMs) and Vision Language Models (VLMs) have shown impressive reasoning abilities, yet they struggle with spatial understanding and layout consistency when performing fine-grained visual editing. We introduce a Structured Reasoning framework that performs text-conditioned spatial layout editing via scene-graph reasoning. Given an input

J. Scholtz, F. D'Eugenio, R. Maiolino, M. Brazzini

It has been claimed that a fraction of the so-called Little Red Dots (LRDs) are characterised by exponential broad line profiles, which have been ascribed to broadening from electron scattering by an ionised cocoon. In this work, we investigate the H$\alpha$ broad line profiles of 32 AGN, including Little Red Dots (LRDs), Little Blue Dots (LBDs), and X-ray d

Alexandra Zelenin, Alexandra Zhuravlyova

Weight-Decomposed Low-Rank Adaptation (DoRA) extends LoRA by decoupling weight magnitude from direction, but its forward pass requires the row-wise norm of W + sBA, a computation that every major framework we surveyed implements by materializing the dense [d_out, d_in] product BA. At d_in = 8192 and rank r = 384, a single module's norm requires about 512 MB

Wooseok Jang, Seonghu Jeon, Jisang Han, Jinhyeok Choi

While recent advances in generative latent spaces have driven substantial progress in single-image generation, the optimal latent space for novel view synthesis (NVS) remains largely unexplored. In particular, NVS requires geometrically consistent generation across viewpoints, but existing approaches typically operate in a view-independent VAE latent space.

Yu Liu, Ailun Wang, Yu Xia, Zhi Wang

Absolute binding free energy (ABFE) calculations offer a theoretically rigorous approach for predicting protein--ligand binding affinities without the scaffold constraints of relative binding free energy (RBFE) perturbations. However, broad adoption of ABFE in high-throughput hit discovery campaigns has been hindered by high computational costs and a lack of

Konstantinos Dareiotis, Máté Gerencsér

We show that the It\^o solutions of the nonlinear stochastic heat equation $$ \partial_t u^\varepsilon- \Delta u^\varepsilon =\varepsilon^{3/4} g (u^\varepsilon) \nabla \xi_\varepsilon, $$ where $ \xi_\varepsilon$ denotes the mollification in space at scale $\varepsilon>0$ of a space-time white noise $\xi$, converge in law, as $\varepsilon\to 0$, to the solu

Zhengyao Lv, Menghan Xia, Xintao Wang, Kwan-Yee K. Wong

Diffusion-based video super-resolution (VSR) has recently achieved remarkable fidelity but still suffers from prohibitive sampling costs. While distribution matching distillation (DMD) can accelerate diffusion models toward one-step generation, directly applying it to VSR often results in training instability alongside degraded and insufficient supervision.

Yixuan Luo, Feng Qiao, Zhexiao Xiong, Yanjing Li

Optical flow estimation is a fundamental problem in computer vision, yet the reliance on expensive ground-truth annotations limits the scalability of supervised approaches. Although unsupervised and semi-supervised methods alleviate this issue, they often suffer from unreliable supervision signals based on brightness constancy and smoothness assumptions, lea

Danielle Wampler, Ralf Bundschuh

RNA binding proteins play a crucial role in post-transcriptional gene regulation by controlling the transport, processing, and translation of their target RNAs. Post-transcriptional gene regulation leads to the differential expression of genetic material and loss of regulation or over-regulation relates to a large range of cancers and diseases - many of whic

Carlos Bornes, Chiheb Ben Mahmoud, Volker L. Deringer, Christopher J. Heard

Solid state nuclear magnetic resonance (ss-NMR) is one of the most sensitive and popular techniques for structure elucidation in geometrically complex crystalline materials, such as zeolites. Synergistic support from computational modelling is vital to interpret experimental spectra, and relate ss-NMR to atomistic models. Nevertheless, computational predicti

Jaeyun Moon

Microscopic understanding of liquid properties is essential for advancing a wide range of applications from energy applications such as nuclear reactors and batteries to biomedical applications including drug delivery and microfluidics. However, intrinsic dynamic disorder and lack of structural periodicity in liquids have presented fundamental challenges in

Chenchen Zhu, Saksham Suri, Cijo Jose, Maxime Oquab

Running AI models on smart edge devices can unlock versatile user experiences, but presents challenges due to limited compute and the need to handle multiple tasks simultaneously. This requires a vision encoder with small size but powerful and versatile representations. We present our method, Efficient Universal Perception Encoder (EUPE), which offers both i

Gu Zhang, Qicheng Xu, Haozhe Zhang, Jianhan Ma

Dexterous manipulation remains challenging due to the cost of collecting real-robot teleoperation data, the heterogeneity of hand embodiments, and the high dimensionality of control. We present UniDex, a robot foundation suite that couples a large-scale robot-centric dataset with a unified vision-language-action (VLA) policy and a practical human-data captur

Hung-Chieh Fang, Amber Xie, Jennifer Grannen, Kenneth Llontop

Performing in-hand, contact-rich, and long-horizon dexterous manipulation remains an unsolved challenge in robotics. Prior hand dexterity works have considered each of these three challenges in isolation, yet do not combine these skills into a single, complex task. To further test the capabilities of dexterity, we propose drumming as a testbed for dexterous

Håvard Bakke Bjerkevik, Joseph Dorfer, Linda Kleist, Torsten Ueckerdt

We consider the problem of reconfiguring non-crossing spanning trees on point sets. For a set $P$ of $n$ points in general position in the plane, the flip graph $F(P)$ has a vertex for each non-crossing spanning tree on $P$ and an edge between any two spanning trees that can be transformed into each other by the exchange of a single edge. This flip graph has

Lang Cheng, Xiaobo Guo, Yuhan Li, Jun Tao

Exceptional points (EPs) in quasinormal mode (QNM) spectra are non-Hermitian degeneracies at which both the eigenvalues and eigenfunctions coalesce. In this paper, we identify an EP in the scalar QNM spectrum of hairy black holes in the Einstein-Maxwell-scalar theory by scanning the parameter space. We then investigate its implications for ringdown signals b

Carolin Holtermann, Minh Duc Bui, Kaitlyn Zhou, Valentin Hofmann

Hundreds of millions of people rely on large language models (LLMs) for education, work, and even healthcare. Yet these models are known to reproduce and amplify social biases present in their training data. Moreover, text-based interfaces remain a barrier for many, for example, users with limited literacy, motor impairments, or mobile-only devices. Voice in

Abhisha Garg, Suraj Srivastava, Varsha Dubey, Aditya Jagannatham

We develop a pragmatic multi-user (MU) massive multiple-input multiple-output (MIMO) channel model tailored to the THz band, encompassing factors such as molecular absorption, reflection losses and multipath diffused ray components. Next, we propose a novel semi-blind based channel state information (CSI) acquisition technique i.e. MU whitening decorrelation

Michael Messere, Kirill Tchernyshyov, Mary E. Putman, Greg L. Bryan

This paper explores the extent to which the circumgalactic medium (CGM) of Milky Way-like galaxies is located in an extended, ionized, disklike structure. To test this hypothesis, we analyze the spatial and kinematic distributions of different ion species within a sample of MW-like systems in IllustrisTNG. We model commonly observed ions (HI, MgII, SiIV, CIV

Nathan Chapelier-Laget, Thomas Gerber, Nicolas Jacon, Cédric Lecouvey

We introduce and study the notion of entropy of affine permutations and prove that it coincides with the atomic length associated with the sum of the fundamental weights for a type $A$ affine root system, as defined by the first two authors. We then establish an analogue of the Granville-Ono theorem by showing that any nonnegative integer can be realised as

Baiying Liu, Chi-Heng Lo, Brian Wen

In this paper, we explicitly classify the corank 4 unitary representations of symplectic or split odd special orthogonal groups over non-Archimedean local fields of characteristic zero, by classifying Arthur representations of corank 4 and verifying the corresponding unitary dual conjecture recently proposed by Hazeltine-Jiang-Liu-Lo-Zhang in [HJLLZ24].

Claudia Islas-Vargas, L. Ricardo Montoya, Carlos A. Vital-José, Oliver T. Unke

Sodium-ion batteries require anodes that combine high capacity, low operating voltage, fast Na-ion transport, and mechanical stability, which conventional anodes struggle to deliver. Here, we use the SpookyNet machine-learning force field (MLFF) together with all-electron density-functional theory calculations to characterize Na storage in aminobenzene-funct

Atta Ur Rehman Sherwani, Emma Pearce, Philipp Hildenstein, Felix Mauerhoff

Plastic pollution has become a critical global challenge, with microplastics pervading ecosystems and entering human food chains. Effectively monitoring this widespread contamination demands rapid, reliable, and portable material identification techniques that often elude conventional Raman and FTIR spectroscopy. Undetected photon spectroscopy within a nonli

Lucas H. Ueda, João G. T. Lima, Pedro R. Corrêa, Flávio O. Simões

This paper presents SelfTTS, a text-to-speech (TTS) model designed for cross-speaker style transfer that eliminates the need for external pre-trained speaker or emotion encoders. The architecture achieves emotional expressivity in neutral speakers through an explicit disentanglement strategy utilizing Gradient Reversal Layers (GRL) combined with cosine simil

Jayesh Badwaik, Mathis Bode, Michal Rajski, Andreas Herten

The increasing heterogeneity of high-performance computing (HPC) systems and the transition to exascale architectures require systematic and reproducible performance evaluation across diverse workloads. While continuous integration (CI) ensures functional correctness in software engineering, performance and energy efficiency in HPC are typically evaluated ou

Tali Pinsky, Federico Salmoiraghi

We present a new approach to hyperbolic plugs, via a construction of bicontact plugs on 3-manifolds with boundary that are surface bundles over the circle. The boundary components are quasi transverse tori, and we prove a gluing theorem that allows us to produce closed manifolds carrying new transitive Anosov flows. We show that a toroidal manifold produced

Jeffri Murrugarra-Llerena, Pranav Chitale, Zicheng Liu, Kai Ao

Social group detection, or the identification of humans involved in reciprocal interpersonal interactions (e.g., family members, friends, and customers and merchants), is a crucial component of social intelligence needed for agents transacting in the world. The few existing benchmarks for social group detection are limited by low scene diversity and reliance

Changxiao Cai, Gen Li

Diffusion language models (DLMs) have emerged as a promising alternative to autoregressive (AR) models for language modeling, allowing flexible generation order and parallel generation of multiple tokens. However, this flexibility introduces a challenge absent in AR models: the \emph{decoding strategy} -- which determines the order and number of tokens gener

Ling Yue, Kushal Raj Bhandari, Ching-Yun Ko, Dhaval Patel

Large language model (LLM)-based systems are becoming increasingly popular for solving tasks by constructing executable workflows that interleave LLM calls, information retrieval, tool use, code execution, memory updates, and verification. This survey reviews recent methods for designing and optimizing such workflows, which we treat as agentic computation gr

Nicole Reindl, David Jones, Todd Hillwig, Marcelo M. Miller Bertolami

Close binary central stars of PNe offer a unique window for investigating the conditions immediately following the ejection of a common envelope (CE). Double eclipsing and double-lined double systems are particularly valuable as they provide minimally model-dependent constraints on fundamental binary parameters. We report that the nucleus of Pa13 (P=0.3988d)

Anson Kost, Huaiyu Duan

We apply the Landau damping theory to collective flavor excitations in dense neutrino media. We demonstrate explicitly that the (meta-)stable modes of the collective flavor excitations in a physical dense neutrino gas are indeed weakly damped. We show that there should in general exist infinitely many collective flavor excitation modes most of which are damp