arXiv preprints from January 1, 2026 through September 26, 2026 — 00:08:21 EST
Accurate and automated lesion segmentation in Positron Emission Tomography / Computed Tomography (PET/CT) imaging is essential for cancer diagnosis and therapy planning. This paper presents a Swin Transformer UNet 3D (SwinUNet3D) framework for lesion segmentation in Fluorodeoxyglucose Positron Emission Tomography / Computed Tomography...
We investigate the Cauchy-Dirichlet problem for linear parabolic equations in divergence form. Under mild assumptions on the source term and the domain, we prove the existence of globally Hölder continuous solutions. Notably, our results accommodate data exhibiting singularities nearly as critical as the inverse square of the distance...
The bulk-disclination correspondence (BDC) is a fundamental concept in Hermitian systems that has been widely applied to predict disclination states. Recently, disclination states have also been observed and experimentally verified in non-Hermitian systems with C6 lattice symmetry, where gain and loss are introduced to induce...
We present photometric properties of 161 galaxy candidates at $z=7-14$ selected from the second data release (DR2) of BEACON, a JWST Cycle 2 pure-parallel NIRCam imaging program. Carefully selected from 36 independent pointings (corresponding to $\sim350$\,arcmin$^2$ sky coverage), and hence with reduced cosmic variance, our galaxy...
We prove an upper bound for the Morse index of min-max stationary integral varifolds realizing the $d$-dimensional $p$-width of a closed Riemannian manifold.
For a given second order elliptic operation $\mathcal{L}$ in a domain $Ω\subset{\mathbb{R}}^\mathbf{N}$, $\mathbf{N}\ $, and a compact set $\mathbf{K}\subsetΩ$, order $\mathbf{N}$-$2$-Ahlfors-David regular, we define the space $\mathcal{H}^{\mathbf{r}+ω}_{\mathcal{L}}(\mathbf{K})$ of continuous functions $f(x),\, x\in\mathbf{K}$,...
Socio-demographic prompting (SDP) - prompting Large Language Models (LLMs) using demographic proxies to generate culturally aligned outputs - often shows LLM responses as stereotypical and biased. While effective in assessing LLMs' cultural competency, SDP is prone to confounding factors such as prompt sensitivity, decoding...
Humanoid robots are envisioned as general-purpose platforms in human-centered environments, yet their deployment is limited by vulnerability to falls and the risks posed by rigid metal-plastic structures to people and surroundings. We introduce a soft-rigid co-design framework that leverages non-Newtonian fluid-based soft responsive...
Precise day-ahead forecasts for electricity prices are crucial to ensure efficient portfolio management, support strategic decision-making for power plant operations, enable efficient battery storage optimization, and facilitate demand response planning. However, developing an accurate prediction model is highly challenging in an...
Linear queries, as the basis of broad analysis tasks, are often released through privacy mechanisms based on differential privacy (DP), the most popular framework for privacy protection. However, DP adopts a context-free definition that operates independently of the data-generating distribution. In this paper, we revisit the privacy...
As an agent-level reasoning and coordination paradigm, Multi-Agent Debate (MAD) orchestrates multiple agents through structured debate to improve answer quality and support complex reasoning. However, existing research on MAD suffers from two fundamental limitations: evaluations are conducted under fragmented and inconsistent...
Using a modified curve shortening flow, we construct $λ$-Angenent curve, which was first constructed by the shooting method.
We present two novel generative geometric deep learning frameworks, termed Flow Matching PointNet and Diffusion PointNet, for predicting fluid flow variables on irregular geometries by incorporating PointNet into flow matching and diffusion models, respectively. In these frameworks, a reverse generative process reconstructs physical...
We study traces of Hecke operators on spaces of elliptic cusp forms and Drinfeld cusp forms and show that, modulo any prime power, these traces are periodic in the weight.
On-policy reinforcement learning (RL), particularly Proximal Policy Optimization (PPO) and Group Relative Policy Optimization (GRPO), has become the dominant paradigm for fine-tuning large language models (LLMs). While policy ratio clipping stabilizes training, this heuristic hard constraint incurs a fundamental cost: it...
Recently, the Dark Energy Spectroscopic Instrument Data Release 2 (DESI DR2) suggests that the dark energy in our universe might be evolving, favoring the Chevallier-Polarski-Linder (CPL) parameterization and a lower Hubble constant. In our previous work, it has been reported that cosmological model with the non-singular bounce...
Preference alignment methods such as RLHF and Direct Preference Optimization (DPO) improve instruction following, but they can also reinforce hallucinations when preference judgments reward fluency and confidence over factual correctness. We introduce F-DPO (Factuality-aware Direct Preference Optimization), a simple extension of DPO...
In the universal framework of simplified $t$-channel dark matter models, the calculation of the relic abundance can be dominated by mediator annihilation when the dark matter and mediator masses are almost degenerate. We analyze four representative models with scalar and fermionic mediators, confront them with direct detection limits...
Long-context question answering (QA) over literary texts poses significant challenges for modern large language models, particularly in low-resource languages. We address the scarcity of long-context QA resources for Indic languages by introducing LittiChoQA, the largest literary QA dataset to date covering many languages spoken in...
We propose Self-Augmented Residual 3D Gaussian Splatting (SA-ResGS), a novel framework to stabilize uncertainty quantification and enhancing uncertainty-aware supervision in next-best-view (NBV) selection for active scene reconstruction. SA-ResGS improves both the reliability of uncertainty estimates and their effectiveness for...
Medical conversational AI (AI) plays a pivotal role in the development of safer and more effective medical dialogue systems. However, existing benchmarks and evaluation frameworks for assessing the information-gathering and diagnostic reasoning abilities of medical large language models (LLMs) have not been rigorously evaluated. To...
A photorealistic and controllable 3D caricaturization framework for faces is introduced. We start with an intrinsic Gaussian curvature-based surface exaggeration technique, which, when coupled with texture, tends to produce over-smoothed renders. To address this, we resort to 3D Gaussian Splatting (3DGS), which has recently been shown...
AlYN and AlScN have recently emerged as promising nitride materials that can be integrated with GaN to form two-dimensional electron gases (2DEGs) at heterojunctions. Electron transport properties in these heterostructures have been enhanced through careful design and optimization of epitaxial growth conditions. In this work, we...
The structural evolution with temperature of pure silica (SiO2), sodium-silicate (5Na2O-95SiO2, 10Na2O-90SiO2 and 25Na2O-75SiO2) and albite (15Na2O-15Al2O3-75SiO2) glasses previously densified from hot compression is monitored with a combination of small and wide angle x-ray scattering. Transient scattering maxima in the length scales...
The regular expression matching problem asks whether a given regular expression of length $m$ matches a given string of length $n$. As is well known, the problem can be solved in $O(nm)$ time using Thompson's algorithm. Moreover, recent studies have shown that regular expression matching extended with a practical extension called...