arXiv preprints from January 1, 2026 through September 30, 2026 — 15:56:48 EST
Talking head generation creates lifelike avatars from static portraits for virtual communication and content creation. However, current models do not yet convey the feeling of truly interactive communication, often generating one-way responses that lack emotional engagement. We identify two key challenges toward truly interactive...
We aim to evaluate how well the variation of small-scale magnetic fields on the stellar surface can be monitored with time-series observations. Further, we aim to establish to what extent the measured total unsigned magnetic field traces other activity indicators. We measured the total unsigned magnetic field on four young, stars...
We construct in detail the irreducible representations of the BMS group with super rotations in three and four dimensions that have the same rest frame momenta as the massive and massless Poincare point particles. We compare these representations to those of the Poincare group and also to the analogous representations of global BMS....
The adoption of Centralized Radio Access Network (C-RAN) architectures requires fronthaul systems capable of carrying large volumes of radio data while meeting stringent delay and jitter requirements. Ethernet Passive Optical Networks (EPONs) have emerged as a promising fronthaul solution due to their cost efficiency and compatibility...
In this article, we study the mixed fourth moments of Hecke--Maass cusp forms and Eisenstein series with type $(2, 2)$. Under the assumptions of the Generalized Riemann Hypothesis (GRH) and the Generalized Ramanujan Conjecture (GRC), we establish asymptotic formulas for these moments. Our results give an interesting...
Despite the rapid success of Large Vision-Language Models (LVLMs), a persistent challenge is their tendency to generate hallucinated content, undermining reliability in real-world use. Existing training-free methods address hallucinations but face two limitations: (i) they rely on narrow assumptions about hallucination sources, and...
Reliable building height estimation is essential for various urban applications. Spaceborne SAR tomography (TomoSAR) provides weather-independent, side-looking observations that capture facade-level structure, offering a promising alternative to conventional optical methods. However, TomoSAR point clouds often suffer from noise,...
We study the nonleptonic decays of $Ω_{b}\rightarrowΩ_{c} P (V)$ with eight pseudoscalar and vector mesons using the naive factorization approach. We analyze all relevant topologies (the tree-level, color-suppressed, and penguin) of these decays and calculate the decay amplitude for each separately. We determine the decay rates,...
Background: Understanding electronic interactions in protein active sites is fundamental to drug discovery and enzyme engineering, but remains computationally challenging due to exponential scaling of quantum mechanical calculations. Results: We present a quantum-classical hybrid framework for simulating protein fragment electronic...
This paper introduces Interpretability-Guided Bi-objective Optimization (IGBO), a framework that trains interpretable models by incorporating structured domain knowledge via a bi-objective formulation. IGBO encodes feature importance hierarchies as a Directed Acyclic Graph (DAG) via Central Limit Theorem-based construction and uses...
We prove $\ell^p(\mathbb{Z}^n)-$estimates for long $r$-variational seminorm of two families of averages: discrete Birch-Magyar averages, for $r>max\{p,p'\}$ with $p>\frac{2c_{\mathfrak{R}}-2}{2c_{\mathfrak{R}}-3}$ and discrete Hardy-Littlewood type averages over certain algebraic varieties, for $r>max\{p,p'\}$ with $p>1$. Further, we...
We study a reputational cheap-talk environment in which a judge, who is privately and imperfectly informed about a state, must choose between two speakers of unknown reliability. Exactly one speaker is an expert who perfectly observes the state, while the other is a quack with no information. Both speakers seek to be selected, while...
We establish the global gradient bounds for weak solutions to the elliptic variational inequality with two-sided obstructions, associated with a $p(x)$-Laplacian type operator involving degenerate or singular matrix weights. Under the optimal regularity assumptions on the matrix-valued weight, suitable geometric flatness of the...
We report new experimental constraints on the generalized version of the gravity-induced decoherence model originally proposed by Károlyházy. Using data collected by the VIP Collaboration at the INFN Gran Sasso National Laboratory with a high-purity germanium detector, we derive an improved lower bound on the spatial correlation...
The Hubble tension is commonly framed as a discrepancy between local, late-time measurements favoring $H_0 \approx 73$ km s$^{-1}$ Mpc$^{-1}$ and early-time, Sound-Horizon-based measurements favoring $H_0 \approx 67$ km s$^{-1}$ Mpc$^{-1}$. We challenge this viewpoint by analyzing 88 Sound Horizon Free $H_0$ measurements, categorized...
Solar-pumped lasers, predominantly based on neodymium gain media, offer a promising route to renewable laser-energy conversion and space-based photonics; however, their performance has been constrained by thermal loading and limited power scalability. Here, we propose and numerically investigate a solar-pumped ytterbium thin-disk gain...
This paper establishes Lipschitz stability for the simultaneous recovery of a variable density coefficient and the initial displacement in a damped biharmonic wave equation. The data consist of the boundary Cauchy data for the Laplacian of the solution, \(Δu |_{\partial Ω}\) and \( \partial_{n}(Δu)|_{\partial Ω}.\) We first prove that...
Large Protein Language Models have shown strong potential for generative protein design, yet they frequently produce structural hallucinations, generating sequences with high linguistic likelihood that fold into thermodynamically unstable conformations. Existing alignment approaches such as Direct Preference Optimization are limited...
In this paper, we provide some sufficient conditions for the compactness of weighted composition operators on Dirichlet space. Furthermore, we characterize the numerical range of certain classes of weighted composition operators on Dirichlet space and establish the criteria that guarantee the inclusion of zero within the numerical...
Image-based deep learning provides a non-invasive, scalable solution for monitoring potato quality during storage, addressing key challenges such as sprout detection, weight loss estimation, and shelf-life prediction. In this study, images and corresponding weight data were collected over a 200-day period under controlled temperature...
Deploying large language models (LLMs) in mobile and edge computing environments is constrained by limited on-device resources, scarce wireless bandwidth, and frequent model evolution. Although edge-cloud collaborative inference with speculative decoding (SD) can reduce end-to-end latency by executing a lightweight draft model at the...
The dynamics of filaments in flow are central to understanding a wide range of biological and soft-matter systems, yet their behavior under time-dependent forcing remains poorly understood. Here, we investigate the long-time dynamics of Brownian inextensible elastic filaments subjected to strong uniform oscillatory shear by combining...
Differential equations with state-dependent delays define a semiflow of continuously differentiable solution operators in general only on the associated {\it solution manifold} $X\subset C^1([-h,0],\mathbb{R}^n)$. For systems with discrete state-dependent delays we construct a diffeomorphism on a neighbourhood of $X$ which takes $X$...
Large language models (LLMs) frequently produce contextual hallucinations, where generated content contradicts or ignores information explicitly stated in the prompt. Such errors are particularly problematic in deterministic automation workflows, where inputs are fixed and correctness is unambiguous. We introduce a simple and...
Large Language Models (LLMs) are integral to modern AI applications, but their safety alignment mechanisms can be bypassed through adversarial prompt engineering. This study investigates emoji-based jailbreaking, where emoji sequences are embedded in textual prompts to trigger harmful and unethical outputs from LLMs. We evaluated 50...