March 2026 arXiv papers — page 22

Edward J. Yoon

We present ITQ3_S (Interleaved Ternary Quantization -- Specialized), a novel 3-bit weight quantization format for LLMs integrating TurboQuant (TQ), a rotation-domain strategy based on the Fast Walsh-Hadamard Transform (FWHT). Conventional 3-bit methods suffer precision loss from heavy-tailed weight distributions and inter-channel outliers. ITQ3_S pre-rotates

Alexander S. Petty

For a prime p and base b, the collision invariant $S_{\ell}(p)$, introduced in the companion paper, is a function of $p \bmod b^{\ell+1}$ and therefore lives on the finite group $(\mathbb{Z}/b^{\ell+1}\mathbb{Z})^{\times}$. Its Fourier expansion over Dirichlet characters modulo $b^{\ell+1}$ is the collision transform. The reflection identity forces all even-

Susana Furtado, Charles Johnson

Here we give a procedure to construct a reciprocal matrix for which the right and entrywise inverse left Perron eigenvectors have any pair of given orders. An explicit example when the matrix is of size 4 is presented. In particular, it gives an afirmative answer to the question posed in a recent manuscript by Bozóki and Csató (2026) about the existence of a

Sai Keerthana Karnam, Abhirup Kundu, Jashn Arora, Manish Jain

Social media serves as a primary source of information in the current digital era. Many people consume a vast range of information in a very short span, yet, amidst the stream of genuine information, fake news and rumors continue to spread. The need for effective detection models is becoming increasingly critical. Past user behavior and user engagement on a

Ansgar Pernice

We study an extension of the quantum linear Boltzmann equation describing irreversible momentum-space dynamics of an open quantum system under strong continuous monitoring. The monitored observable is taken to be a quadratic form in an extended, purely Euclidean four-dimensional momentum space, without assuming any fixed signature at the microscopic level. I

Ben Godfrey, Doug Dawson

The development of administrative and computational research tools requires reliable programmatic interfaces with the cluster scheduler. The Research Computing and Data (RCD) team at Clemson University has developed the Palmetto API, a proxy for the native Slurm RESTful interface, slurmrestd, while providing advanced authentication, authorization, and cachin

Bharath Krishnamurthy, Ajita Rattani

Recent multimodal face generation models address the spatial control limitations of text-to-image diffusion models by augmenting text-based conditioning with spatial priors such as segmentation masks, sketches, or edge maps. This multimodal fusion enables controllable synthesis aligned with both high-level semantic intent and low-level structural layout. How

Abhishek Saraswat, Weiyue Zhou, Nayoung Kim, Jaron F. Cota

This paper presents results from systematic investigations conducted in the HYPERION facility to quantify permeabilities of hydrogen isotopes in FLiBe over a temperature range of 773K - 973K. To address the knowledge gap resulting from widely scattered transport parameters reported in the earlier studies, HYPERION experiments incorporate specific provisions

Keven Alkhoury

Filled rubber-like materials are widely used in engineering applications and are well known to exhibit the Mullins effect. In this work, an established local constitutive model from the literature is extended to a non-local setting to resolve the mesh dependence inherent to the local approach. Non-local effects are incorporated using two separate approaches:

Cristian Fanjul, A. A. Aligia

We incorporate spin-orbit coupling (SOC) into effective Kugel-Khomskii models for the $n=1$ and $n=2$ members of the Ruddlesden-Popper series Sr$_{n+1}$Cr$_n$O$_{3n+1}$. These model contain interacting spins 1 and pseudospins 1/2 at each site describing spin and orbitals degrees of freedom respectively. We solve the models at zero temperature using pseudospi

Andreas Feuerpfeil, Atanu Maity, Ronny Thomale, Yasir Iqbal

We investigate Higgs criticality in candidate U(1) Dirac spin liquids across a family of depleted triangular lattices: the triangular, kagome, and maple-leaf geometries. For each, we identify the symmetry-allowed spinon-pairing channel connecting the U(1) state to a proximate $\mathbb{Z}_2$ spin liquid, deriving the corresponding quantum electrodynamics (QED

Ana M. Conrado, Rubén García-Benito, Rosa M. González Delgado, Bahar Bidaran

The mass in the Universe is distributed non-uniformly, originating the Large Scale Structure (LSS), characterised by clusters, filaments, walls and voids. Galaxies in voids are bluer, later type, less massive, and have slower evolution than galaxies in denser environments. The effect of the void environment on properties such as star formation rate (SFR) is

Xueying Zheng, Gabriele Ponti, Nicola Locatelli, Beate Stelzer

F, G, K and M type stars are the most abundant stellar population in the Milky Way and are expected to contribute to its diffuse X-ray emission. Yet their intrinsic average X-ray spectrum remains poorly constrained due to their faint X-ray luminosities, leaving their collective role in the X-ray background of the Milky Way uncertain. We analysed the volume-c

Aclécio J. Santos, Jean C. Pereira, Guilherme V. Raffo

This paper proposes an interior-point framework for constrained optimization problems whose decision variables evolve on matrix Lie groups. The proposed method, termed the Matrix Lie Group Interior-Point Method (MLG-IPM), operates directly on the group structure using a minimal Lie algebra parametrization, avoiding redundant matrix representations and elimin

P. Cristofari, G. Emery, T. Lubrano di Vavaria, H. Costantini

SNR G106.3+2.7 is a complex TeV emitting source whose emission is still poorly understood. It has especially been at the center of numerous discussions on its potential for being a supernova remnant (SNR) PeVatron, since its gamma-ray spectra seems not to exhibit any significant suppression in the multi--TeV range, up to $\sim 600$ TeV, thereby indicating th

Ivan Dornic, Robert Conte

We determine the full persistence probability distribution for a non-Markovian stochastic process, motivated by first-passage questions arising in interacting spin systems and allied systems. We show that this distribution is governed by a distinguished Painlevé VI system arising from an exact Fredholm Pfaffian structure associated with the integrable sech k

Jiabao Peng, Longjie Zhang, Zhenxu Bai, Stephan Fritzsche

Highly coherent lasers are central to modern photonics. To date, high-coherence operation has been achieved predominantly in microcavity and fiber-based platforms. More recently, free-space Brillouin-laser experiments have revealed unusually strong noise suppression whose physical origin cannot be explained by conventional continuous-medium models developed

Jonathan Ramírez, Santiago Esteban Perez Bergliaffa

We establish a correspondence between higher-derivative gravitational scalar-tensor theories of the form $Ψ(R,(\nabla R)^2,\Box R)$ and generalized hybrid metric-Palatini models $f(R,\mathcal{R})$. Restricting to the physically relevant case of linear dependence on $\Box R$, we make explicit that both frameworks can be reformulated in the Einstein frame as G

Chiara Badiali, Rafael Almeida, Thales Silva, Jorge Vieira

Velocity-controlled spatio-temporal (ST) laser drivers offer a route to tailoring laser-plasma interactions by allowing the velocity of the intensity peak to be controlled independently of the envelope group velocity. In this work, we present a simulation-design workflow for PIC modelling of subluminal velocity-controlled ST pulses in OSIRIS based on a Maxwe

Jonas K König, Jamie M Fitzgerald, Daniel Erkensten, Ermin Malic

Microscopic insights into nonlinear interactions are essential for advancing polaritonic devices. Existing studies often rely on phenomenological models that overlook important many-body processes. Based on a material-specific and predictive approach, we investigate monolayer and homobilayer MoS$_2$ embedded in a Fabry-Pérot cavity to characterize the exchan

Atharva Pandit, Fred C. MacKintosh, Abhinav Sharma

Disordered fiber networks exhibit a floppy to rigid mechanical phase transition as a function of connectivity. Sub-isostatically connected networks can undergo this transition via straining. Critical exponents governing this transition have been estimated theoretically and by numerical simulations of various types of networks. In this study, we present impro

Vedantha Srinivas Kasturi, Juan Nieves, Federico Sánchez, Joanna Ewa Sobczyk

We explore the combined lepton and hadron kinematic observables from the exclusive Valencia $2p2h$ model. We present variables of interest which are available due to the exclusive kinematics and compare them with the democratically distributed outgoing nucleon kinematics as currently treated in neutrino event generators. We also show the effect of nuclear re

Tomohiro Tanaka, Yusuke Nishida

We investigate two-cluster scattering in a quasi-one-dimensional Bose gas. We focus on the effective three-body interaction induced by transverse confinement, which is the leading term for breaking integrability in the quasi-one-dimensional setting. Exploiting the Lüscher formula and the integrability of the Lieb-Liniger Bose gas, we find a finite and positi

Liangzu Liu, Yiyan Wang, Yinjun Wu, Runze Su

Cardinality estimation (CardEst) still remains a challenging problem for DBMS. Recent years have witnessed the success of ML-based cardinality estimators in outperforming traditional methods. However, these solutions suffer from poor generalizability to new data or query distribution, inability to handle complex queries, and substantial data preparation over

Michael R. R. Good, Eric V. Linder

We present an analytic flat-spacetime accelerating boundary analog of Hawking-type emission that possesses infinite asymptotic acceleration (and radial acceleration in the black hole analog) but finite total radiated energy (and zero surface gravity in the black hole analog). We perform a unified study of its scattering symmetry, horizon formation, asymptoti

Edward Wijaya

Deep learning models for drug-like molecules and proteins overwhelmingly reuse transformer architectures designed for natural language, yet whether molecular sequences benefit from different designs has not been systematically tested. We deploy autonomous architecture search via an agent across three sequence types (SMILES, protein, and English text as contr

Norichika Sago, Ryuichi Fujita, Soichiro Isoyama, Hiroyuki Nakano

We analytically derive the secular changes of the orbital parameters, i.e., energy, angular momentum, and Carter constant, for general bound orbits in Kerr spacetime, at leading order in the mass ratio, through the 6th post-Newtonian (6PN) order and the 16th order in orbital eccentricity. We validate the formulas against high-precision numerical Teukolsky re

Max Disselnmeyer, Thomas Bömer, Laura Dörr, Bastian Amberg

Buffer zones are essential in production systems to decouple sequential processes. In dense floor storage environments, such as space-constrained brownfield facilities, manual operation is increasingly challenged by severe labor shortages and rising operational costs. Automating these zones requires solving the Buffer Storage, Retrieval, and Reshuffling Prob

BESIII Collaboration, M. Ablikim, M. N. Achasov, P. Adlarson

An amplitude analysis of the singly Cabibbo-suppressed decay $D^0 \to K^+ K^- π^0 π^0$ is performed, for the first time, to determine the relative magnitudes and phases of different intermediate processes. The analysis uses $e^+e^-$ collision data collected with the BESIII detector at the center-of-mass energy 3.773~GeV corresponding to an integrated luminos

Marc Seidel, Martina Maggio, Frank Allgöwer

Deadline misses are more common in real-world systems than one may expect. The weakly-hard task model has become a standard abstraction to describe and analyze how often these misses occur, and has been especially used in control applications. Most existing control approaches check whether a controller manages to stabilize the system it controls when its imp

Sebastian Jaroszewicz, Nahuel Mendez, Maria P. Beccar-Varela, Maria Cristina Mariani

Multifractal Detrended Fluctuation Analysis (MFDFA) has emerged as a standard tool for characterizing scale invariance in complex systems, yet its application to discrete spin models is frequently marred by reports of ``spurious multifractality'' that contradict established theory. In this work, we resolve this controversy by establishing a rigorous

Shao-Zhen Lin, Sham Tlili, Jean-François Rupprecht

We present a rotationally invariant viscous vertex model that accounts for both cortical and bulk dissipation of cells. The vanishing substrate-friction limit is enforced via Lagrange multipliers, which also provides a framework for implementing various boundary conditions, such as fixed boundaries and prescribed tractions. Building on this formulation, we i

Mo Xiong, Jize Han, Chuanzhen Cao, Jinbin Li

The scalable preparation of large photon-number (Fock) states is a long-standing frontier in quantum science, with direct implications for quantum metrology and bosonic quantum information processing. Despite substantial progress at small photon numbers, extending state generation to large photon numbers while maintaining high fidelity and operating determin

F. Alemanno, Q. An, P. Azzarello, F. C. T. Barbato

The DArk Matter Particle Explorer (DAMPE) is a space-borne high-energy particle detector that surveys the $γ$-ray sky above$\sim 2~\rm GeV$ with a peak acceptance of $\sim 0.2~\rm m^2\,sr$. With the 102 months of data collected by DAMPE, we show that the Fermi bubbles are detected at a significance of $\sim 26σ$ and identify a GeV excess in the direction of

Teodor Iličin, Rok Žitko

We investigate the superconducting Anderson impurity model for interacting quantum dot Josephson junctions with spin-orbit coupling and a term accounting for tunnelling through higher-energy orbitals. These elements establish the conditions required for spin polarization in the absence of external magnetic field at finite superconducting phase bias. This Ham

J. Choppin de Janvry, S. Gontcho A Gontcho, U. Seljak, A. Baleato Lizancos

The calibration of tomographic redshift distributions is essential for cosmological analysis of weak lensing data. In this work, we calibrate all four tomographic bins of the Hyper Suprime Camera (HSC) weak lensing catalog with the Dark Energy Spectroscopic Instrument (DESI) Data Release 1 and 2 using the clustering redshifts technique. We include z > 1.2 re

Timothy De Deyn, Pat Lank, Kabeer Manali-Rahul, Sridhar Venkatesh

This work is concerned with categorical methods for studying singularities. Our focus is on birational derived splinters, which is a notion that extends the definition of rational singularities beyond varieties over fields of characteristic zero. Particularly, we show that an invariant called `level' in the associated derived category measures the failur

Ashim Dahal, Ankit Ghimire, Saydul Akbar Murad, Nick Rahimi

Long-video multimodal question answering requires structured reasoning over visual evidence and dialogue, but Large Vision-Language Models (LVLMs) are constrained by context-window and compute limits. We propose POVQA, which compresses each second into a temporally pooled image (1 fps pooled images) to maintain dense temporal coverage under a fixed token bud

Petru Jebelean, Jean Mawhin, Calin Serban

We are concerned with the existence of $T$-periodic solutions to an equation of type $$\left (|u'(t))|^{p(t)-2} u'(t) \right )'+f(u(t))u'(t)+g(u(t))=h(t)\quad \mbox{ in }[0,T]$$ where $p:[0,T]\to(1,\infty)$ with $p(0)=p(T)$ and $h$ are continuous on $[0,T]$, $f,g$ are also continuous on $[0,\infty)$, respectively $(0,\infty)$. The mapping $g$

Sonia Petrini, Antoni Casas-i-Muñoz, Jordi Cluet-i-Martinell, Mengxue Wang

Zipf's law of abbreviation, namely the tendency of more frequent words to be shorter, has been viewed as a manifestation of compression, i.e. the minimization of the length of forms -- a universal principle of natural communication. Although the claim that languages are optimized has become trendy, attempts to measure the degree of optimization of langua

Dexing Huang, Shiao Wang, Fan Zhang, Xiao Wang

Robust visual object tracking (VOT) remains challenging in high-speed motion scenarios, where conventional RGB sensors suffer from severe motion blur and performance degradation. Event cameras, with microsecond temporal resolution and high dynamic range, provide complementary structural cues that can potentially compensate for these limitations. However, exi

Andrew W. Singletary, Max H. Cohen, Tamas G. Molnar, Aaron D. Ames

The advancement of autonomous systems -- from legged robots to self-driving vehicles and aircraft -- necessitates executing increasingly high-performance and dynamic motions without ever putting the system or its environment in harm's way. In this paper, we introduce Guardrails -- a novel runtime assurance mechanism that guarantees dynamic safety for autonom

Michael Petrov, Boo Carmans, Josef Soucek, Akhil Kuriakose

This paper aims to extend the understanding of the mechanism of photo-electrical detection of magnetic resonance (PDMR) in nitrogen-vacancy (NV) centres. This technique is particularly important for development of solid-state quantum computing platforms. In particular, we report on the new insight in the photocurrent (PC) generation and charge cycling in the

Sungyong Chung, Yanlin Zhang, Nachuan Li, Dana Monzer

Car-following behavior is fundamental to traffic flow theory, yet traditional models often fail to capture the stochasticity of naturalistic driving. This paper introduces a new car-following modeling category called the empirical probabilistic paradigm, which bypasses conventional parametric assumptions. Within this paradigm, we propose the Markov Chain Car

Beleicia Bullock, James A. Landay, Michael S. Bernstein

Metaphors enable designers to communicate their ideal user experience for platforms. Yet, we often do not know if these design metaphors match users' actual experiences. In this work, we compare design and user metaphors across three different platforms: ChatGPT, Twitter, and YouTube. We build on prior methods to elicit 554 user metaphors, as well as ratings

Geoffrey W. Marcy

Artificial intelligence is projected to increase U.S. data centre power demand beyond 100 gigawatt by 2035 and global demand toward 1 terrawatt. In response, companies and governments have proposed placing computing infrastructure in sun-synchronous low-Earth orbit, where continuous sunlight could supply electrical power. Generating 5 GW would require solar

Tomas Berggren, Nedialko Bradinoff

We study a family of periodically weighted Aztec diamond dimer models near their turning points. We establish that, asymptotically, as $N\rightarrow\infty$, their fluctuations there, scaled by $\sqrt{N}$, are described by a marked GUE-corners process. This limiting point process is constructed by assigning a Bernoulli mark independently to each particle in a

Christopher Cruz

We present ATLAS-RTC, a runtime control system for autoregressive language models that enforces structured output during decoding. ATLAS-RTC monitors generation at each step, detects drift from output contracts using lightweight signals, and applies targeted interventions such as biasing, masking, and rollback. Unlike post-hoc validation or static constraine

Haokun Zhou

Stereo needs features that preserve fine cross view correspondence rather than only semantic similarity. Recent self supervised vision models transfer well, but they are not built for this goal, and geometry focused methods often rely on a binocular decoder or another explicit linkage module during pretraining. BINO asks whether strong binocular structure ca

Matthew Loftus

We prove that the persistence diagram of the sublevel set filtration of the quadratic form f(x) = x^T M x restricted to the unit sphere S^{n-1} is analytically determined by the eigenvalues of the symmetric matrix M. By Morse theory, the diagram has exactly n-1 finite bars, with the k-th bar living in homological dimension k-1 and having length equal to the

Yuda Li, Xiang Yin

This paper investigates one-step backward reachability for uncertain max-plus linear systems with additive disturbances. Given a target set, the problem is to compute the set of states from which there exists an admissible control input such that, for all admissible disturbances, the successor state remains in the target set. This problem is closely related

Petr Vlachopulos

We develop a bicategorical framework for entropy and Landauer's principle in which entropy-ordered state spaces are treated not merely through deterministic monotone maps, but through open many-to-many interfaces encoding feasible realizations between logical and thermodynamic descriptions. This leads to the bicategory of open entropy systems, whose objects

Patrizio Cintioli

We construct a nonrecursive set \(A\le_T\emptyset'\) and a uniformly computable family of sets \(C_0,C_1,\dots\), all bounded finite-one equivalent to \(A\), such that the corresponding \(1\)-degrees form a copy of the dense linear order \((\mathbb Q,\le)\). Motivated by a recent preprint of Richter, Stephan, and Zhang, which shows that bounded finite-one de

Wei-Yuan Su, Ruijie Zhang, Zheng Zhang

Vision Transformers (ViTs) achieve state-of-the-art performance but suffer from the $O(N^2)$ complexity of self-attention, making inference costly for high-resolution inputs. To address this bottleneck, token pruning has emerged as a critical technique to accelerate inference. Most existing methods rely on the [CLS] token to estimate patch importance. Howeve

Ioannis Kousek, Vicente Saavedra-Araya

Given a semigroup $G$ and a bounded function $f: G \to \mathbb{C}$, a topological Furstenberg system of $f$ is a topological dynamical system $\mathbb{X}=(X, (T_g)_{g \in G})$ that encodes the dynamical behaviour of $f$. We show that $\mathbb{X}$ is unique up to topological isomorphism, thus providing a topological analogue of the measurable case established

Tripti Shukla, Zsolt Kira

Large vision-language models (VLMs) frequently suffer from hallucinations, generating content that is inconsistent with visual inputs. Existing methods typically address this problem through post-hoc filtering, additional training objectives, or external verification, but they do not intervene during the decoding process when hallucinations arise. In this wo

Zhang XueGuang

Totally similar physical process in tidal disruption events (TDEs) basically indicates that there should be potential parameter to distinguish variability properties of TDEs from the other transient events having different physical processes. Here, we try to report such a parameter, the timescale ratio $R_{2/1,rd}$ of rise timescale $t_{1/2,r}$ (from half-ma

Keeryn Johnson, Muhammad Ahmed, Charlie Lang, Sahib Thethi

This paper investigates how the integration of large language models influences gameplay, playability, and player experience in game development. We report a collaborative autoethnographic study of two game projects in which LLMs were embedded as architectural components. Reflective narratives and development artifacts were analyzed using gameplay, playabili

Lucas Valenca, Ronnie de Souza Santos

Background. As digital technologies increasingly shape social domains such as healthcare, public safety, entertainment, and education, software engineering has engaged with ethical and political concerns primarily through the notion of algorithmic fairness. Aim. This study challenges the limits of software engineering approaches to fairness by analyzing how

Francesca Albertini, Domenico D'Alessandro

In this paper, we demonstrate an approach to quantum robust control based on the tools of geometric optimal control. The central objects of interest are the sensitivity functions defined as the coefficients in the Taylor expansion of the trajectory with respect to the (unknown, small) parameters which describe the deviation of the actual model from nominal o

Yunda Yan, Chenxi Tao, Jinya Su, Cunjia Liu

Ensuring both safety and stability remains a fundamental challenge in learning-based control, where goal-oriented policies often neglect system constraints and closed-loop state convergence. To address this limitation, this paper introduces the Predictive Safety--Stability Filter (PS2F), a unified predictive filter framework that guarantees constraint satisf

Subrata Ghosh, Rosalin Mohanty, Yuwei Sun, Soumi Mondal

Two-dimensional (2D) van der Waals (vdW) multiferroics have emerged as a promising platform for next-generation multifunctional devices. Although recent studies have demonstrated that artificial heterostructures can combine dual ferroic orders and exhibit strong magnetoelectric coupling, their performance is sometimes limited by poor interface quality and in

Irene Kim, Sai Tanmay Reddy Chakkera, Alexandros Graikos, Dimitris Samaras

Monocular surface normal estimators trained on large-scale RGB-normal data often perform poorly in the edge cases of reflective, textureless, and dark surfaces. Polarization encodes surface orientation independently of texture and albedo, offering a physics-based complement for these cases. Existing polarization methods, however, require multi-view capture o

Thomas Bernert, Rob Sullivan, Jeroen Winkel, Shujie Yang

We show the simplicity of the automorphism groups of the generic $n$-hypertournament and the semigeneric tournament, and determine the normal subgroups of the automorphism groups of several other ultrahomogeneous oriented graphs. We also give a new proof of the simplicity of the automorphism group of the dense $\frac{2\pi}{n}$-local order $\mathbb{S}(n)$ for

Matteo Guida, Yulia Otmakhova, Eduard Hovy, Lea Frermann

Framing theory posits that how information is presented shapes audience responses, but computational work has largely ignored audience reactions. While recent work showed that article framing systematically shapes the content of reader responses, this paper asks: does framing also affect response quality? Analyzing 1M comments across 2.7K news articles, we o

Tao Su, Baiting Xie, Chenglong Yu

We propose a log-concavity conjecture for BPS invariants arising in the enumerative geometry of planar curve singularities, identified with the local Euler obstructions of Severi strata in their versal deformations. We further extend this conjecture to ruling polynomials of Legendrian links and to E-polynomials of character varieties. We establish these conj

Vsevolod Biryukov, Kamal Choudhary, Timur Bazhirov

Artificial intelligence (AI) and machine learning (ML) models in materials science are predominantly trained on ideal bulk crystals, limiting their transferability to real-world applications where surfaces, interfaces, and defects dominate. We present Mat3ra-2D, an open-source framework for the rapid design of realistic two-dimensional materials and related

Matthew Loftus

We investigate whether spectral properties of neural network weight matrices can predict test accuracy. Under controlled label noise variation, the tail index alpha of the eigenvalue distribution at the network's bottleneck layer predicts test accuracy with leave-one-out R^2 = 0.984 (21 noise levels, 3 seeds per level), far exceeding all baselines: the best

Kihyun Yu, Seoungbin Bae, Dabeen Lee

We study safe reinforcement learning in finite-horizon linear mixture constrained Markov decision processes (CMDPs) with adversarial rewards under full-information feedback and an unknown transition kernel. We propose a primal-dual policy optimization algorithm that achieves regret and constraint violation bounds of $\widetilde{O}(\sqrt{d^2 H^3 K})$ under mi

Christin Pagels, Simon Hacks, Rob Henk Bemthuis

Enterprise Architecture Debt (EA Debt) arises from suboptimal design decisions and misaligned components that can degrade an organization's IT landscape over time. Early indicators, Enterprise Architecture Smells (EA Smells), are currently mainly detected manually or only from structured artifacts, leaving much unstructured documentation under-analyzed. This

Eduardo Brito, Fernando Castillo, Amnir Hadachi, Ulrich Norbisrath

Reliable use of real-world data requires confidence that recorded evidence reflects what actually occurred at the moment of capture. In adversarial or incentive-misaligned cyber-physical settings, device-centric provenance and post-capture verification are insufficient to provide that guarantee. This paper builds on Proof-of-Location (PoL) as a baseline for

Ting Ji, Laura Liu, Yulong Wang, Jiahe Xing

This paper proposes a specification test for the conventional distributional assumptions of error terms in binary choice models, focusing on its tail properties. Based on extreme value theory, we first establish that the tail index of the unobserved error can be recovered by that of the observed covariates. The null hypothesis of the index being zero essenti

Jnaneshwar Das

We propose a variational framework in which the kernel function k : X x X -> R, interpreted as the foundational object encoding what distinctions an agent can represent, is treated as a dynamical variable subject to path entropy maximization (Maximum Caliber, MaxCal). Each kernel defines a representational structure over which an information geometry on prob

Anastasia Boushmelev, Matthew Black, Oliver Witzel

Semileptonic $B_{(s)}$ decays are of great phenomenological interest because they allow to determine e.g. CKM matrix elements or test lepton flavor universality. Taking advantage of already existing lattice data, we demonstrate the analysis steps to extract the four form factors describing exclusive semileptonic $B_s\to D_s^*\ell\nu_\ell$ decays using the na

Yuri Mishin

We present a detailed analysis of the universal relationship between grain boundary (GB) free energy and GB self-diffusion coefficient derived by Borisov et al. (1964). This relationship was expressed by a simple equation that was used in many publications to predict GB energies on the basis of experimental diffusion data. Meanwhile, the physical assumptions

Benno Weck, Pablo Puentes, Andrea Poltronieri, Satyajeet Prabhu

The evaluation of music understanding in Large Audio-Language Models (LALMs) requires a rigorously defined benchmark that truly tests whether models can perceive and interpret music, a standard that current data methodologies frequently fail to meet. This paper introduces a meticulously structured approach to music evaluation, proposing a new dataset of 320

Anis Attiaoui, Shunda Chen, Joseph C. Woicik, J. Zach Lentz

The functional properties of semiconductors are typically controlled by tailoring their chemical composition and their state of strain, and by controlling their long-range structural order, including the presence of extended defects such as dislocations. In addition to these approaches, theoretical predictions suggest that short-range order (SRO) of atoms in

Gregory Magarshak

We introduce probabilistic language tries (PLTs), a unified representation that makes explicit the prefix structure implicitly defined by any generative model over sequences. By assigning to each outgoing edge the conditional probability of the corresponding token or action, a PLT simultaneously serves as: (i) an optimal lossless compressor via frequency-wei

Masoud S. Sakha, Rushikesh Kamalapurkar, Sean Meyn

Relative temporal-difference (TD) learning was introduced to mitigate the slow convergence of TD methods when the discount factor approaches one by subtracting a baseline from the temporal-difference update. While this idea has been studied in the tabular setting, stability guarantees with function approximation remain poorly understood. This paper analyzes

Jarosław Hryszko

Language models cannot be random. This paper introduces Entropic Deviation (ED), the normalised KL divergence between a model's token distribution and the uniform distribution, and measures it systematically across 31,200 generations spanning seven models, two architectures (transformer and state space), nine prompt categories, three temperatures, and five l

Elsayed Elamir

This study develops two robust, quantile-sliced moment systems, mean and median absolute deviation (MAD and MedAD moments), to serve as foundational tools in parametric modeling, statistical inference, and describing distributional location, scale, skewness, and tail behavior in settings where classical moments and L-moments fail. MAD moments use block-wise

Boning Li, Patrick Hautle, Duhan Zhang, Liangping Zhu

Optically addressable molecular triplet spins provide a chemically tunable platform for quantum application, but their coherence is often limited by interactions with surrounding spin baths. Here we demonstrate controlled suppression of nuclear-bath-induced decoherence in photoexcited triplet spins of pentacene co-crystallized in high-purity naphthalene sing

Jeremiah Birrell, Xiaoxi Shen

We study finite-sample statistical performance guarantees for distributionally robust optimization (DRO) with optimal transport (OT) and OT-regularized divergence model neighborhoods. Specifically, we derive concentration inequalities for supervised learning via DRO-based adversarial training, as commonly employed to enhance the adversarial robustness of mac

Mohammad Farhoudi, Hamidreza Mazandarani, Masoud Shokrnezhad, Tarik Taleb

The proliferation of users, devices, and novel vehicular applications - propelled by advancements in autonomous systems and connected technologies - is precipitating an unprecedented surge in novel services. These emerging services require substantial bandwidth allocation, adherence to stringent Quality of Service (QoS) parameters, and energy-efficient imple

Chao Ying, Siyi Deng, Yang Ning, Jiwei Zhao

Semi-supervised learning has attracted significant attention due to the proliferation of applications featuring limited labeled data but abundant unlabeled data. In this paper, we examine the statistical inference problem in an assumption-lean framework which involves a high-dimensional regression parameter, defined by minimizing the least squares, within th

Elchin Suleymanov

Human decision makers increasingly delegate choices to AI agents, raising a natural question: does the AI implement the human principal's preferences or pursue its own? To study this question using revealed preference techniques, I introduce the Luce Alignment Model, where the AI's choices are a mixture of two Luce rules, one reflecting the human's preferenc

Stanisław Domoradzki, Małgorzata Stawiska, Mykhailo Zarichnyi

Inspired by the recent 90th anniversary of the Scottish Book we present some reflections about its impact. First we discuss new areas of mathematics it helped launch. Then we argue that it was actively used in stimulating the interests and results of junior mathematicians and students. Also, we summarize the progress during the decade that has passed since t

Ming Liu, Yunbei Zhang, Shilong Liu, Liwen Wang

Video generation models produce visually coherent content but struggle with tasks requiring spatial reasoning and multi-step planning. Reinforcement learning (RL) offers a path to improve generalization, but its effectiveness in video reasoning hinges on reward design -- a challenge that has received little systematic study. We investigate this problem by ad

Pietro Baldi, Vesa Julin, Domenico Angelo La Manna

We consider the Dirichlet-Neumann operator for a nearly spherical domain in R^n, and prove sharp analytic and tame estimates in Sobolev class. The novelty of this paper concerns technical improvements, the most important of which are the independence of the analyticity radius on the high norms and the regularity loss of one in the elevation function. These p

Khai Nguyen, Yang Ni, Peter Mueller

We review recently proposed Bayesian approaches for clustering high-dimensional data. After identifying the main limitations of available approaches, we introduce an alternative framework based on vertical consensus inference (VCI) to mitigate the curse of dimensionality in high-dimensional Bayesian clustering. VCI builds on the idea of consensus Monte Carlo

Glen MacLachlan, Joseph Creech, Rubeel Muhammad Iqbal, Clark Gaylord

Migrating heterogeneous high-performance computing (HPC) systems to resource-aware scheduling introduces both technical and behavioral challenges, particularly in production environments with established user workflows. This paper presents a case study of transitioning a production academic HPC cluster from node-exclusive to consumable resource scheduling mi

Samin Mahdizadeh Sani, Max Ku, Nima Jamali, Matina Mahdizadeh Sani

Advances in diffusion, autoregressive, and hybrid models have enabled high-quality image synthesis for tasks such as text-to-image, editing, and reference-guided composition. Yet, existing benchmarks remain limited, either focus on isolated tasks, cover only narrow domains, or provide opaque scores without explaining failure modes. We introduce \textbf{Image

Borja Odriozola Schick

Every system that maintains a large language model conversation beyond a single session faces two inescapable constraints: the context window is finite, and information quality degrades with accumulated volume. We formalize these constraints as axioms and derive a single governing principle -- the Root Theorem of Context Engineering: \emph{maximize signal-to

Isabel Barrio Sanchez, Catalin Trenchea, Wenlong Pei

In this paper we study the long-time stability of the Cauchy one-leg theta-methods for the two-dimensional NavierStokes equations. We establish the uniform dissipativity in H^1, in the sense that the semi-discrete-in-time approximations possess a global attractor for a small enough time step, using the discrete Gronwall lemma and the discrete uniform Gronwal

Miao Zhang

Focal mechanisms of small earthquakes are critical for characterizing faults and regional stress. P-wave polarities and S/P amplitude ratios (e.g., in HASH) are widely used to determine focal mechanisms for small earthquakes, but first-motion picking can be difficult for emergent onsets, and S/P ratios are often highly inconsistent because of imperfect veloc

Tyler Burleigh

Automated scoring of student work at scale requires balancing accuracy against cost and latency. In "cascade" systems, small language models (LMs) handle easier scoring tasks while escalating harder ones to larger LMs -- but the challenge is determining which cases to escalate. We explore verbalized confidence -- asking the LM to state a numerical confidence

Rauan Akylzhanov

Large language models fragment Kazakh text into many more tokens than equivalent English text, because their tokenizers were built for high-resource languages. This tokenizer tax inflates compute, shortens the effective context window, and weakens the model's grip on Kazakh morphology. We propose to bypass the tokenizer entirely by feeding raw bytes through

Mirko Amico

Phase kickback is a fundamental primitive that is used in many quantum algorithms, such as quantum phase estimation. Here we observe that by using information about the controlled unitary, we can replace the controlled unitary with an uncontrolled one at the cost of introducing controlled state preparations. We then show how this modified phase kickback can

Alexandros S. Kalafatelis, Nikolaos Nomikos, Vasileios Nikolakakis, Nikolaos Tsoulakos

Smart shipping operations increasingly depend on collaborative AI, yet the underlying data are generated across vessels with uneven connectivity, limited backhaul, and clear commercial sensitivity. In such settings, server-coordinated FL remains a weak systems assumption, depending on a reachable aggregation point and repeated wide-area synchronization, both

Dario Paape

I use the Pythia scaling suite (Biderman et al. 2023) to investigate if and how two well-known polarity illusions, the NPI illusion and the depth charge illusion, arise in LLMs. The NPI illusion becomes weaker and ultimately disappears as model size increases, while the depth charge illusion becomes stronger in larger models. The results have implications fo