April 2026 arXiv papers — page 95

Jianming Tong, Jingtian Dang, Simon Langowski, Tianhao Huang

Zero-knowledge proof (ZKP) provers remain costly because multi-scalar multiplication (MSM) and number-theoretic transforms (NTTs) dominate runtime as they need significant computation. AI ASICs such as TPUs provide massive matrix throughput and SotA energy efficiency. We present MORPH, the first framework that reformulates ZKP kernels to match AI-ASIC execut

Jiakun Zheng, Ting Xiao, Shiqin Cao, Xinran Li

Text-to-motion (T2M) generation aims to control the behavior of a target character via textual descriptions. Leveraging text-motion paired datasets, existing T2M models have achieved impressive performance in generating high-quality motions within the distribution of their training data. However, their performance deteriorates notably when the motion descrip

Yuki Okamura, Ren Yatsunami, Kumiko Kameishi, Oliver Posani

(1)Cross-border data transfers have become a matter of daily occurrence against the backdrop of the development of cloud computing and artificial intelligence. Consequently, where a data leak gives rise to civil liability, the determination of that liability inevitably assumes an international dimension involving foreign elements. (2)As is starkly demonstrat

Farid Diaf, Alex Moriani, Rym Smaï, Graham Andrew Smith

In this paper, we study the class of Weil--Petersson circle homeomorphisms from the point of view of three-dimensional anti-de Sitter space $\mathbf{AdS}^{2,1}$. We show that a homeomorphism $\varphi:\mathbf{RP}^1\to\mathbf{RP}^1$ is Weil--Petersson if and only if its graph, viewed as a curve in the boundary at infinity of $\mathbf{AdS}^{2,1}$, is the asympt

Prasoon Goyal, Sattvik Sahai, Michael Johnston, Hangjie Shi

Post-training Large Language Models requires diverse, high-quality data which is rare and costly to obtain, especially in low resource domains and for multi-turn conversations. Common solutions are crowdsourcing or synthetic generation, but both often yield low-quality or low-diversity data. We introduce Adversarial Arena for building high quality conversati

Dahua Gao, Ruichao Liu, Minxi Yang, Shuai Ma

Generative Semantic Communication (GSC) is a promising solution for image transmission over narrow-band and high-noise channels. However, existing GSC methods rely on long, indirect transport trajectories from a Gaussian to an image distribution guided by semantics, causing severe hallucination and high computational cost. To address this, we propose a gener

Pufan Li, Bi'an Du, Shenghe Zheng, Junyi Yao

Text-driven 3D scene editing has recently attracted increasing attention. Most existing methods follow a render-edit-optimize pipeline, where multi-view images are rendered from a 3D scene, edited with 2D image editors, and then used to optimize the underlying 3D representation. However, cross-view inconsistency remains a major bottleneck. Although recent me

Tuan Van Vo, Tan Q. Nguyen, Khang Nguyen, Nhat Xuan Tran

Vision-Language-Action (VLA) models have gained much attention from the research community thanks to their strength in translating multimodal observations with linguistic instructions into desired robotic actions. Despite their advancements, VLAs often overlook explicit reasoning and learn the functional input-action mappings, omitting crucial logical steps,

Yuanmin Liu

We prove the semistable reduction theorem for $\mathcal{E}^{\dag}_K$-valued and $K$-valued overconvergent $F$-isocrystals over $k((t))$-varieties which were introduced by Lazda and P\'{a}l. As an application, we prove the finite dimensionality of $\mathcal{E}^{\dag}_K$-valued rigid cohomology with compact support.

Shavkat Ayupov, Abdireymov Arislanbay, Bakhtiyor Yusupov

In this work, we describe local and 2-local $\frac12$-derivations of infinite-dimensional Lie algebras. We prove that all local and 2-local $\frac12$-derivations of the Witt algebra as well as of the positive Witt algebra and the classical one-sided Witt algebra are $\frac12$-derivations. We also give an example of an infinite-dimensional Lie algebra with a

Miaojing Shi, Jun Huang, Zijie Yue, Hanli Wang

Referring video object segmentation (RVOS) aims to segment the target instance in a video, referred by a text expression. Conventional approaches are mostly supervised learning, requiring expensive pixel-level mask annotations. To tackle it, weakly-supervised RVOS has recently been proposed to replace mask annotations with bounding boxes or points, which are

Sanchari Das, Dhiman Goswami, Michelle Melo, Aditya Johri

As digital systems increasingly rely on pervasive data collection and inference, educating future designers and researchers about Usable Privacy has become a critical need for HCI. However, privacy education in higher education is often fragmented, theory-heavy, or detached from real-world applications. Thus, in this paper, we present the design, implementat

You Zhou, Peter E. Driscoll, Mingming Zhang, Christian Reinhardt

Accretional heating of Earth's interior during formation is pivotal to its subsequent thermal and chemical evolution. In particular, impact heating of Earth's core is expected, but its amplitude and radial distribution within the core is unknown and could influence the onset of the geodynamo. The uncertainty is due, in part, to the lack of constraints on the

Bui The Trung, Do Minh Duc, Nguyen Van Vinh, Bui Nguyen Quoc Trinh

The democratization of ubiquitous AI hinges on deploying sophisticated reasoning capabilities on resource-constrained devices. However, Small Language Models (SLMs) often face a "reasoning gap", particularly in non-English languages like Vietnamese, where they struggle to maintain coherent chains of thought. This paper investigates Test-Time Scaling strategi

Peter Lebedev-Stepanov

An analytical solution is obtained for the problem of the slow movement of a small drop of a fluid in another immiscible fluid in an infinitely large reservoir with the boundary condition of the normal slip and/or tangential partial slip at the interface. That generalizes the conventional Navier and Maxwellian boundary conditions of partial slip. Normal slip

Ziqi Guo

The goal of our work is to construct a class of morphisms between two canonical line bundles on integral models of PEL Shimura varieties via Kodaira--Spencer maps, and explicitly compute such morphisms and their effects on the canonical metrics of line bundles. This result provides a concrete method for comparing two canonical line bundles and the correspond

Xuhui Zhang, Chunjie Wang, Wenchao Liu, Huijun Xing

The accelerating advancement of intelligent transportation systems has established consumer-oriented vehicular networks (CVNs) as a critical infrastructure for next-generation connected mobility. However, the high mobility of vehicular users (VUs) introduces significant channel state information (CSI) uncertainty, which severely undermines the performance of

K. Uzawa

Sub-barrier photo-induced fission in $^{236}$U$(\gamma,f)$ is investigated within the non-equilibrium Green function (NEGF) method. A model space for the fission process is constructed by superposing Skyrme-Hartree-Fock wave functions along the fission path allowing the particle-hole excitation. Then, the transition from the photo-absorption channel to the f

Haokun Lin, Xinle Jia, Haobo Xu, Bingchen Yao

The MXFP4 microscaling format, which partitions tensors into blocks of 32 elements sharing an E8M0 scaling factor, has emerged as a promising substrate for efficient LLM inference, backed by native hardware support on NVIDIA Blackwell Tensor Cores. However, activation outliers pose a unique challenge under this format: a single outlier inflates the shared bl

Dhiman Goswami, Al Nahian Bin Emran, Md Hasan Ullah Sadi, Sanchari Das

Online propaganda detection pipelines expose measurable privacy risks at multiple stages including data collection, feature extraction, and model inference. We conduct a structured analysis of $162$ peer-reviewed studies and formalize the problem using the Propaganda Risk Online Mitigation and Privacy-preserving Tactics (PROMPT) framework. PROMPT models risk

Domenic P. J. Germano, Federico Frascoli, Robyn P. Araujo, Peter P. Lee

The coordination of the immune system and its components is essential for the body to maintain a healthy status. Recent clinical studies show that breast cancer patients with high Dendritic cell clustering in tumour draining lymph nodes have improved survival outcomes, compared to those with a lower degree of clustering. These results suggest that a specific

Seunghee Koh, Sunghyun Baek, Youngdong Kim, Junmo Kim

Unlearning in large language models (LLMs) has emerged as a promising safeguard against adversarial behaviors. When the forgetting loss is applied uniformly without considering token-level semantic importance, model utility can be unnecessarily degraded. Recent studies have explored token-wise loss regularizers that prioritize informative tokens, but largely

Sichen Ding, Zhiwu Li

Classical opacity theory for discrete-event systems relies strictly on observable event sequences, fundamentally failing to capture security breaches in hybrid architectures where an attacker exploits both classical traces and localized quantum correlations. To address this gap, we formalize current-state opacity within the framework of safe partially observ

Naoya Hatano, Ryota Kawasumi, Hiroki Saito, Hitoshi Tanaka

We show some simple sufficient conditions for which the multilinear embedding theorem holds for fractional sparse operators. By verifying these conditions, we establish the theorem for power weights. We also provide Morrey-type sufficient conditions for which the $L^p \to L^q$, $1<p,q<\infty$, infinitesimal relative bounds hold for Schr\"{o}dinger operators

Lin Jiang, Qingshan She, Jiale Xu, Haiqi Xu

Decoding visual content from electroencephalography (EEG) is important for understanding neural visual representations and developing non-invasive brain-computer interfaces. Existing approaches mainly improve EEG representation learning and cross-modal alignment while treating pretrained visual representations as fixed supervision targets. However, pretraine

Boqun Huang, Yancheng Wang, Wei Guo, Zhaojie Guo

Low-altitude communication networks (LACNs) serve as the critical infrastructure of the emerging low-altitude economy (LAE), supporting services such as drone delivery and infrastructure inspection. However, LACNs operate in highly dynamic three-dimensional (3D) environments characterized by high mobility and predominantly line-of-sight (LoS) propagation, cr

Satoru Utsunomiya, Masaru Isonuma, Junichiro Mori, Ichiro Sakata

As generative AI faces intensifying legal challenges, the machine learning community has increasingly relied on post-hoc mitigation -- especially machine unlearning and inference-time guardrails -- to argue for compliance. This paper argues that such post-hoc mitigation methods cannot retroactively cure liability from unlawful acquisition and training, becau

Pranshav Gajjar, Vijay K Shah

Large language models (LLMs) are increasingly deployed in the telecommunications domain for critical tasks, relying heavily on Retrieval-Augmented Generation (RAG) to adapt general-purpose models to continuously evolving standards. However, a significant gap exists in evaluating the embedding models that power these RAG pipelines, as general-purpose benchmar

Arindam Biswas

We give a uniform explicit construction of finite two-generator presentations for the special linear groups over the integers in all ranks at least three. The construction builds on the generating-pair work of Conder--Liversidge--Vsemirnov and on a standard Tietze-elimination observation pointed out by Button. It recovers Trott's odd-rank generating pair and

Yutian Pang, Daniel Delahaye, John-Paul Clarke

We present a trajectory-based optimization framework for arrival sequencing and scheduling in the terminal maneuvering area (TMA). Unlike node-link scheduling models that reduce trajectories to time-delay variables, the proposed method computes implementable per-aircraft speed profiles and path extensions that achieve required landing separation through term

Sebastian Pardo-Guerra, Anil Thapa, Jonathan Washburn

Let $G$ be a finite connected simple graph with a chosen orientation of its edges. For the edge potential $\psi(t)=\cosh t-1,$ we minimize $\sum_{e\in E^\to}\psi(z_e)$ over each affine class $\omega+dC^0(G)\subset C^1(G)$. The minimizer is the unique representative satisfying the nonlinear coclosed equation $\delta\sinh z=0,$ and hence defines a nonlinear se

Yingtao Tian

LLM agents in markets present algorithmic collusion risks. While prior work shows LLM agents reach supracompetitive prices through tacit coordination, existing research focuses on hand-crafted prompts. The emerging paradigm of prompt optimization necessitates new methodologies for understanding autonomous agent behavior. We investigate whether prompt optimiz

Hongxu Jiang, Fei Li, Boxiao Yu, Ying Zhang

Three-dimensional (3D) medical image enhancement, including denoising and super-resolution, is critical for clinical diagnosis in CT, PET, and MRI. Although diffusion models have shown remarkable success in 2D medical imaging, scaling them to high-resolution 3D volumes remains computationally prohibitive due to lengthy diffusion trajectories over high-dimens

Mohammadtaher Safarzadeh, Hitesh Laxmichand Patel, Afshin Orojlooyjadid, Graham Horwood

Large language models (LLMs) have achieved strong performance on natural language to SQL (NL2SQL) benchmarks, yet their reported accuracy may be inflated by contamination from benchmark queries or structurally similar patterns seen during training. We introduce SPENCE (Syntactic Probing and Evaluation of NL2SQL Contamination Effects), a controlled syntactic

Pranshav Gajjar, Manan Tiwari, Sayanta Seth, Vijay K. Shah

Data scarcity remains a fundamental bottleneck in applying deep learning to wireless communication problems, particularly in scenarios where collecting labeled Radio Frequency (RF) data is expensive, time-consuming, or operationally constrained. This paper proposes LLM-AUG, a data augmentation framework that leverages in-context learning in large language mo

Yuan Fang, Yiming Luo, Aimin Zhou, Fei Tan

Ensuring the safety of large language models (LLMs) requires robust red teaming, yet the systematic synthesis of high-quality toxic data remains under-explored. We propose Reverse Constitutional AI (R-CAI), a framework for automated and controllable adversarial data generation that moves beyond isolated jailbreak prompts. By inverting a harmless constitution

Xiaohan Zou, Roshan Sridhar, Mohammadtaher Safarzadeh, Dan Roth

The reliability of VLM-as-a-Judge is critical for the automatic evaluation of vision-language models (VLMs). Despite recent progress, our analysis reveals that VLM-as-a-Judge often pays limited attention to the image when making decisions. Instead, they often blindly favor the more informative answer, even when they can recognize it conflicts with the image

Tai Hyun Yoon

Interference is conventionally attributed to path-accumulated phase differences, with measurement treated as a passive readout. Here we demonstrate that single-particle interference is governed by the relative phase between the prepared quantum state and the detector-defined measurement basis -- a joint quantity that is operationally inaccessible in any conv

Yutaro Enomoto, Subaru Shibai, Kiwamu Izumi

Space-borne gravitational-wave telescopes are key to extend the observation band below $10\,\mathrm{Hz}$. The use of inter-satellite optical cavities linked by heterodyne interferometry is a promising approach to reach the sensitivity level of $10^{-22}/\sqrt{\mathrm{Hz}}$ in the decihertz band. While heterodyne interferometry is advantageous for relaxing ar

Shuyuan Yang, Jinchuan Hou, Kan He

Employing mutually-commuting von Neumann algebras to represent the algebra of observables on quantum systems provides a framework for studying quantum information theory in systems with infinite degrees of freedom and quantum field theory, yielding many profound results that differ from non-relativistic quantum systems. In this paper, we establish a mutually

Mohammed Kharma, Ahmed Sabbah, Radi Jarrar, Samer Zain

This paper presents a controlled quasi-experimental developer study examining whether a layer-based security training package is associated with improved security quality in LLM-assisted implementation of an identity-centric Java Spring Boot backend. The study uses a mixed design with a within-subject pre-training versus post-training comparison and an explo

Jingxin Yan, Lin Liu, Oliver Dukes, Qizhai Li

Discussion on ``Regression by Composition'' by Farewell, Daniel, Stensrud, and Huitfeldt

Rongyuan Tan, Jue Zhang, Zhuozhao Li, Qingwei Lin

Interpretability tools are increasingly used to analyze failures of Large Language Models (LLMs), yet prior work largely focuses on short prompts or toy settings, leaving their behavior on commonly used benchmarks underexplored. To address this gap, we study contrastive, LRP-based attribution as a practical tool for analyzing LLM failures in realistic settin

Ruixuan Zhao, Oliver Dukes, Linbo Wang, Lin Liu

Discussion on "Regression by Composition" by Farewell, Daniel, Stensrud, and Huitfeldt.

Man-Chun Lee

A quantitative version of the scalar lower bound under $C^0$ convergence was conjectured by Gromov. More recently, Mazurowski and Yao proved that a refined form of Gromov's conjecture holds in dimension three. Furthermore, they constructed examples demonstrating that such a refinement is necessary. In this paper, we establish that the refined quantitative bo

Zuoyu Wu, Yougang Wang, Shihong Liao

Isolated dwarf galaxies are intrinsically faint and difficult to detect. The limited sample size makes it challenging to observationally constrain the physical mechanisms that quench their star formation. To disentangle the quenching mechanisms of isolated dwarfs, we identify a non-negligible population of such galaxies in the TNG50 simulation. In addition t

Hongrui Ma, Huaiqing Zuo

In this paper, we use Hilbert-Samuel multiplicity, Hilbert-Kunz multiplicity, and s-multiplicity to establish a sharp upper bound for the quotient of the generalized Milnor numbers and the Tjurina numbers for isolated hypersurface singularities of any dimension in positive characteristic. Using this result, we also derive an upper bound for the quotient of t

Sanaz Sadat Hosseini, Mona Azarbayjani, Mohammad Pourhomayoun, Hamed Tabkhi

Climate-driven wildfires are intensifying, particularly in urban regions such as Southern California. Yet, traditional fire risk communication tools often fail to gain public trust due to inaccessible design, non-transparent outputs, and limited contextual relevance. These challenges are especially critical in high-risk communities, where trust depends on ho

Abdul Rahman

We extend the Hodge atoms framework of Katzarkov--Kontsevich--Pantev--Yu to one-parameter conifold degenerations of Calabi--Yau threefolds. For a degeneration $\pi\colon X \to \Delta$ whose central fiber $X_0$ has $r$ ordinary double points, we construct a canonical rigid-flexible decomposition of the Hodge atoms of the nearby smooth fiber attached to the co

Anda Cao, Zhuo Gou, Yi Wang, Kaixuan Chen

Merging multiple Low-Rank Adaptation (LoRA) experts into a single backbone is a promising approach for efficient multi-task deployment. While existing methods strive to alleviate interference via weight interpolation or subspace alignment, they rest upon the implicit assumption that all LoRA matrices contribute constructively to the merged model. In this pap

Yali Zhang, Guidong Liu, Shuhuang Xiang

Based on the Hilb-type formula and van der Corput-type lemmas, we present optimal asymptotic estimates for the decay of the Laguerre and Hermite coefficients for functions with algebraic and logarithmic singularities, which in turn yield the convergence rates of the corresponding spectral orthogonal projections. Numerous examples are provided to verify the o

Lixian Chen, Jianhong Tan

Adapting foundation models under resource budgets relies heavily on Parameter-Efficient Fine-Tuning (PEFT), with LoRA being a standard modular solution. However, LoRA suffers from spectral interference. Low-rank updates often concentrate energy on the leading singular directions of pretrained weights, perturbing general capabilities and causing catastrophic

Yihao Zou, Tianming Zheng, Futai Zou, Yue Wu

Fuzzing has become a widely adopted technique for vulnerability discovery, yet it remains ineffective for structured-input programs due to strict syntactic constraints and limited semantic awareness. Traditional greybox fuzzers rely on mutation-based strategies and coarse-grained coverage feedback, which often fail to generate valid inputs and explore deep e

Mengmeng Ge, Takashi Isobe, Xu Jia, Yanan Sun

Understanding physical transformation processes is crucial for both human cognition and artificial intelligence systems, particularly from an egocentric perspective, which serves as a key bridge between humans and machines in action modeling. We define this modeling process as Egocentric Instructed Visual State Transition (EIVST), which involves generating i

Song Tang, Yunxiang Bai, Wenxin Su, Mao Ye

Source-Free Domain Adaptation (SFDA) seeks to adapt a source model, which is pre-trained on a supervised source domain, for a target domain, with only access to unlabeled target training data. Relying on pseudo labeling and/or auxiliary supervision, conventional methods are inevitably error-prone. To mitigate this limitation, in this work we for the first ti

Deyi Wang, Qining Zhang, Lei Ying

This paper considers reinforcement learning from human feedback in a federated learning setting with resource-constrained agents, such as edge devices. We propose an efficient federated RLHF algorithm, named Partitioned, Sign-based Stochastic Zeroth-order Policy Optimization (Par-S$^2$ZPO). The algorithm is built on zeroth-order optimization with binary pert

Brian H. Y. Cheng, Donniel C. Cruz, Otto A. Hannuksela, Davendra S. Hassan

The cosmological principle asserts that the Universe is homogeneous and isotropic on large enough scales. However, alternative cosmological models can bring about anisotropies through local inhomogeneities, anisotropic evolution, or exotic physics. In addition, select studies have also hinted at mild evidence of anisotropies in SNe Ia, CMB, and GRB data, tho

Hanhua Hong, Yizhi LI, Jiaoyan Chen, Sophia Ananiadou

Recent advances in large language models have highlighted their potential to automate computational research, particularly reproducing experimental results. However, existing approaches still use fixed sequential agent pipelines with weak global coordination, which limits their robustness and overall performance. In this work, we propose Hierarchical Researc

Wu Yue

Continuous-control reinforcement learning (RL) often exhibits large closed-loop variance, high-frequency control jitter, and sensitivity to disturbance injection. Existing explanations usually emphasize disturbance sources such as action noise, exploration perturbations, or policy nonsmoothness. This letter studies a complementary amplifier-side perspective:

Kazuhiro Horie

This study presents a numerical analysis framework for pursuit--evasion differential games under the Stackelberg equilibrium. The Semi-DCNLP method is introduced as an optimization solver for flight path optimization problems formulated under the Stackelberg equilibrium. In a spacecraft pursuit-evasion differential game, the Semi-DCNLP method produces trajec

Hangkun Hu, Jingyi Wang, Minggang Wang

The algebraic degree of a network game measures the complexity of its totally mixed Nash equilibria. For sparse multilinear network games, Datta's formula expresses this degree combinatorially in terms of a permanent, but the geometric origin of this formula has remained unclear. In this paper, we provide a tropical-geometric derivation of Datta's formula by

Tim F. Weiss, Alberto Peruzzo

The long-range transmission of quantum information relies on multiple interfaces between photons, acting as flying qubits, and localized memories, serving as repeaters, to mitigate transmission losses. Efficient, long-range transmission necessitates the use of short, picosecond-scale photons, which are markedly different from the narrowband, nanosecond-scale

Chenming Tang, Hsiu-Yuan Huang, Weijie Liu, Junqiang Zheng

Reinforcement learning (RL) has become a prevalent paradigm for training tool calling agents, which typically requires online interactive environments. Existing approaches either rely on training data with ground truth annotations or require advanced proprietary language models (LMs) to synthesize environments that keep fixed once created. In this work, we p

Zhaohua Liang, Zhilin Wang, Renjie Cao, Yining Zhang

Candidate sourcing for recruiters is best viewed as a two-stage retrieval and reranking pipeline with recall as the primary objective under a limited review budget. An upstream production retriever first returns a candidate shortlist for each job description (JD), and our goal is to rerank that shortlist so that qualified candidates appear as high as possibl

Wenjie Diao, Haoqian Pan, Chunxing Yan

Gompf-Scharlemann-Thompson and Meier-Zupan constructed an infinite family of R-links that are potential counterexamples of the generalized property R conjecture. Their works also show that whether these links are stably handleslide trivial is an interesting open problem related to the Slice-Ribbon conjecture. In this work, we implement an algorithm to constr

Haotian Qin, Dongliang Chang, Yueying Gao, Yuexuan Tan

As AI generative models evolve at unprecedented speed, image attribution has become a moving target. New diffusion, adversarial and autoregressive generators appear almost monthly, making existing watermark, classifier and inversion methods obsolete upon release. The core problem lies not in model recognition, but in the inability to adapt attribution itself

Maya Banks, Ritvik Ramkumar

We initiate a study of varieties of minimal degree in weighted projective spaces. We call a weighted projective space $\mathbf{P}(w_0,\dots,w_n)$ divisible if $w_i \mid w_{i+1}$ for all $i$. We provide sharp bounds for when a non-degenerate subvariety of a divisible weighted projective space has minimal degree. We define a weighted notion of $1$-generic matr

Xiaoqi Wu, Xueying Zhan, Wen Li, Junhao Wu

Cryo-electron microscopy (cryo-EM) enables single-particle analysis of biological macromolecules under strict low-dose imaging conditions, but the resulting micrographs often exhibit extremely low signal-to-noise ratios and weak particle visibility. Image denoising is therefore an important preprocessing step for downstream cryo-EM analysis, including partic

Michael Grabchak

We develop the first exact and computationally tractable method for simulating from tempered stable distributions in the infinite variation case, which corresponds to $\alpha\in[1,2)$. A small simulation study shows that the approach works well.

Julieth Saavedra, A. J. Castrillón Vásquez

In this paper, we study the first eigenvalue of the Laplace--Beltrami operator on the Lawson minimal surfaces $\xi_{m,k}$ embedded in the unit three-sphere $\mathbb{S}^3$. Motivated by Yau's conjecture on the first eigenvalue of closed embedded minimal hypersurfaces in the sphere, we develop a symmetry-based approach to the equality $\lambda_1(\xi_{m,k})=2$

Suhyun Lee, Palakorn Achananuparp, Neemesh Yadav, Ee-Peng Lim

Large language models (LLMs) are increasingly explored as scalable tools for mental health counseling, yet evaluating their safety remains challenging due to the interactional and context-dependent nature of clinical harm. Existing evaluation frameworks predominantly assess isolated responses using coarse-grained taxonomies or static datasets, limiting their

Wenqian Zhang

Let $k\geq2$ be an integer. A $k$-tree is a tree with maximum degree at most $k$. In this paper, we give a closure result on spanning $k$-trees of graphs with given minimum degree. Let $\delta\geq1$ be an integer, and $G$ be a connected graph of order $n$ with minimum degree $\delta$. Let $u$ and $v$ be two nonadjacent vertices of $G$ satisfying $d_{G}(u)+d_

Jie Zhang, Jinkun You, Shi Chen, Yicong Zhou

Most existing hyperspectral image super-resolution methods require modifications for different scales, limiting their flexibility in arbitrary-scale reconstruction. 2D Gaussian splatting provides a continuous representation that is compatible with arbitrary-scale super-resolution. Existing methods often rely on rasterization strategies, which may limit flexi

Arya Hadizadeh Moghaddam, Drew Ross, Mohsen Nayebi Kerdabadi, Dongjie Wang

Large Language Models (LLMs) have shown strong promise for mining Electronic Health Records (EHRs) by reasoning over longitudinal clinical information to capture context-rich patient trajectories. However, leveraging LLMs for structured EHRs (e.g., standardized diagnosis and medication codes) presents two key challenges. First, translating time-stamped EHR s

Dong Li, Lei Lu, Jingye Yan, Xinhua Zhao

We traced the origin of very long-periodic pulsations (VLPs) in type-I burst chains on 2024 February 14. Seven successive and repetitive pulsation structures appeared in radio dynamic spectra in the metric waveband, which were simultaneously measured by CBSm, DART, and MUSER-L. A quasi-period at about 160$^{+11}_{-6}$ s, determined by the fast Fourier transf

Tim F. Weiss, Hamed Arianfard, Yang Yang, Alberto Peruzzo

State-of-the-art photon sources based on spontaneous parametric down-conversion (SPDC) currently rely on artificial structuring of the material nonlinearity to satisfy phase-matching conditions. This technique, known as periodic poling, is available only in a limited number of material platforms and introduces additional fabrication steps and errors, which a

Jiayi Tian, Haiduo Huang, Tian Xia, Wenzhe Zhao

We address the challenge of point cloud registration using color information, where traditional methods relying solely on geometric features often struggle in low-overlap and incomplete scenarios. To overcome these limitations, we propose GeGS-PCR, a novel two-stage method that combines geometric, color, and Gaussian information for robust registration. Our

Yuzhe Fu, Hancheng Ye, Cong Guo, Junyao Zhang

Point-based Neural Networks (PNNs) have become a key approach for point cloud processing. However, a core operation in these models, Farthest Point Sampling (FPS), often introduces significant inference latency, especially for large-scale processing. Despite existing CUDA- and hardware-level optimizations, FPS remains a major bottleneck due to exhaustive com

Emine Altuntas, Ian B. Spielman

Experimental access to many-body quantum systems is often limited by measurement backaction, and key dynamical properties are typically obtained by perturbing a system and measuring its response. Here we replace this active paradigm with a minimally invasive protocol based on a pair of weak quantum measurements that leverages measurement backaction as a stre

Hang Yuan, Xiaolin Hu, Yan Wan, Menglin Gao

Text-driven controllable dance generation remains under-explored, primarily due to the severe scarcity of high-quality datasets and the inherent difficulty of articulating complex choreographies. Characterizing dance is particularly challenging owing to its intricate spatial dynamics, strong directionality, and the highly decoupled movements of distinct body

Mehwish Nasim, Sanjeevan Selvaganapathy, Neel Ganapathi Sabhahit, Marie Griesbach

Many benchmarks show that large language models can answer direct questions about culture. We study a different question: do they also change how they speak when culture is only implied by the situation? We evaluate 60 culturally grounded conversational scenarios across five languages in three conditions: a neutral baseline (Prompt A), an explicit cultural i

Muhammad Bilal, Moiz Ali, Mohit Kumar, Fareed Zaffar

Program debloating aims to remove unused code to reduce performance overhead, attack surfaces, and maintenance costs. Over time, debloating has evolved across multiple layers (container, library, and application), each building on the principles of application-level debloating. Despite its central role, application-level debloating continues to rely on imper

Jon-Paul Cacioli

The validity screen (Cacioli, 2026d, 2026e) classifies LLM confidence signals as Valid, Indeterminate, or Invalid. We test whether these classifications predict selective prediction performance. Twenty frontier LLMs from seven families were evaluated on 524 items across six cognitive tracks. Valid models show mean Type 2 AUROC = .624 (SD = .048). Invalid mod

Khang Tran, Khoa Nguyen, Cristian Borcea, NhatHai Phan

Recent advances in large language models for test case generation have improved branch coverage via prompt-engineered mutations. However, they still lack principled mechanisms for steering models toward specific high-risk execution branches, limiting their effectiveness for discovering subtle bugs and security vulnerabilities. We propose GLMTest, the first p

Jon-Paul Cacioli

LLM confidence signals are used for abstention, routing, and safety-critical decisions. No standard practice exists for checking whether a confidence signal carries item-level information before building on it. We transfer the validity screening principle from clinical personality assessment (PAI, MMPI-3) as a portable protocol for benchmark-based LLM confid

Kunyu Zhang, Qiang Li, Vince D. Calhoun, Shujian Yu

Resting-state functional magnetic resonance imaging (fMRI) has emerged as a cornerstone for psychiatric diagnosis, yet most approaches rely on pairwise brain cortical or sub-cortical connectivities that overlooks higher-order interactions (HOIs) central to complex brain dynamics. While hypergraph methods encode HOIs through predefined hyperedges, their const

Masashi W. Kimura, Seong Won Jang, Nisha Geng, Eva Zurek

High-pressure superhydrides have attracted much attention due to their high superconducting critical temperatures ($T_\text{c}$s). Herein, density functional theory (DFT) calculations are used to study the structures and properties, including potential for metal atom disorder and doping-enhanced $T_\text{c}$, within Y-Ca superhydrides with YCaH$_{n}$ ($n=8-2

Jakwang Kim

In this paper, we study the stability of the shadow, a projection of a measure onto the set of couplings with respect to the Wasserstein distance. The shadow was introduced by \citet{Eckstein_Nutz_2022} to analyze the stability of the Sinkhorn algorithm, and was recently revisited by \citet{kim2026extensioncouplingprojectionoptimal} for statistical applicati

Jiangnan Li, Linqing Huang, Xiaowen Yan, Min Gan

Zero-shot learning (ZSL) aims to recognize unseen classes without visual instances. However, existing methods usually assume clean labels, overlooking real-world label noise and ambiguity, which degrades performance. To bridge this gap, we propose the Dynamic Visual-semantic Alignment (DVSA), a robust ZSL framework for learning from ambiguous labels. DVSA us

Jon-Paul Cacioli

Clinical personality assessment screens response validity before interpreting substantive scales. LLM evaluation does not. We apply the validity scaling framework from the PAI and MMPI-3 to metacognitive probe data from 20 frontier models across 524 items. Six validity indices are operationalised: L (maintaining confidence on errors), K (betting on errors),

Haoxiang Jie, Yaoyuan Yan, Xiangyu Wei, Kailin Wang

Visual-Language-Action (VLA) models represent a paradigm shift in embodied AI, yet existing frameworks often struggle with imprecise spatial perception, suboptimal multimodal fusion, and instability in reinforcement learning. To bridge these gaps, we propose OmniVLA-RL, a novel architecture that leverages a Mix-of-Transformers (MoT) design to synergistically

Mamikon S. Ginovyan

In this paper, we survey results on the asymptotic behavior of the variance of the best linear unbiased estimator (BLUE) for the mean of stationary processes. This behavior is influenced by the regularity and memory structures of the observed models. The results show that the asymptotic behavior of the variance of the BLUE is determined solely by the behavio

Mahya Mohammadi, Meryem-Nur Duman, Isa Ahmadalidokht, Mohammad Sadraeian

We investigate quantum spectroscopy with undetected photons for protein detection in the mid-infrared spectral region. Classical Fourier-transform infrared spectroscopy of protein samples (bovine serum albumin and N-terminal pro-brain natriuretic peptide) is used as reference to define the sample's mid-infrared absorption, which is then embedded in a numeric

Bjørn Kjos-Hanssen

Carmo and Jones have presented a sequence of candidate axiom systems for conditional obligation between 1997 and 2022. For their most recent system we demonstrate a limited form of deontic explosion: given that a student does not get the highest possible grade on a test, any other passing grade is acceptable. In addition to that negative result, we give a po

José Francisco Pérez-Barragán

We investigate the generation of standing waves in the model provided by the inhomogeneous telegraph equation under different forcing conditions. We show that sustained standing waves arise only for a specific forcing that is spatially distributed, continuous, and resonant.

Zixuan Liu, Zhiyong Chen, Nan Xue, Shengkang Chen

While distributed device-edge speculative decoding enhances resource utilization across heterogeneous nodes, its performance is often bottlenecked by conventional token-level verification strategies. Such rigid alignment leads to excessive rejections, significantly diminishing the accepted sequence length and increasing interaction rounds under fluctuating w

Niful Islam, Muhammad Anas Raza, Mohammad Wardat

Large Language Models (LLMs) have transformed software development and AI applications. While LLMs are designed for text processing, LLM agents extend this capability by enabling autonomous actions, tool use, and multi-step task completion. As this field grows, developers face new challenges in debugging these complex systems. To address this challenge, we p

Prashant C. Raju

Reliable deployment of language models requires two capabilities that appear distinct but share a common geometric foundation: predicting whether a model will accept targeted behavioral control, and detecting when its internal structure degrades. We show that geometric stability, the consistency of a representation's pairwise distance structure, addresses bo

Luis Bosshart, Max Deter, Leander Heldring, Cathrin Mohr

We digitize and analyze the near-universe of National Socialist German Workers' Party (NSDAP) membership records and link them to newly digitized population and industrial censuses. Four findings emerge. First, as the party expanded, its membership came to resemble the broader population more closely in occupational, demographic, and religious terms. Second,

Xiachong Feng, Deyi Yin, Xiaocheng Feng, Yi Jiang

Games offer a compelling paradigm for developing general reasoning capabilities in language models, as they naturally demand strategic planning, probabilistic inference, and adaptive decision-making. However, existing self-play approaches rely solely on terminal game outcomes, providing no mechanism to distinguish transferable reasoning patterns from game-sp