Papers
Research papers from arXiv and related sources
OpenClaw PRISM: A Zero-Fork, Defense-in-Depth Runtime Security Layer for Tool-Augmented LLM Agents
Tool-augmented LLM agents introduce security risks that extend beyond user-input filtering, including indirect prompt injection through fetched content, unsafe tool execution, credential leakage, a...
Frank Li
Deep Learning-based Assessment of the Relation Between the Third Molar and Mandibular Canal on Panoramic Radiographs using Local, Centralized, and Federated Learning
Impaction of the mandibular third molar in proximity to the mandibular canal increases the risk of inferior alveolar nerve injury. Panoramic radiography is routinely used to assess this relationshi...
Johan Andreas Balle Rubak, Sara Haghighat, Sanyam Jain, Mostafa Aldesoki, Akhilanand Chaurasia, S...
ZeroSense:How Vision matters in Long Context Compression
Recent visual-text compression (VTC) methods, typified by DeepSeek-OCR, report impressive high token compression ratios for long-context modeling tasks by leveraging text-to-image rendering. Howeve...
Yonghan Gao, Zehong Chen, Lijian Xu, Jingzhi Chen, Jingwei Guan, Xingyu Zeng
Acoustic-to-Articulatory Inversion of Clean Speech Using an MRI-Trained Model
Articulatory acoustic inversion reconstructs vocal tract shapes from speech. Real-time magnetic resonance imaging (rt-MRI) allows simultaneous acquisition of both the acoustic speech signal and art...
Sofiane Azzouz, Pierre-André Vuissoz, Yves Laprie
ReDimNet2: Scaling Speaker Verification via Time-Pooled Dimension Reshaping
We present ReDimNet2, an improved neural network architecture for extracting utterance-level speaker representations that builds upon the ReDimNet dimension-reshaping framework. The key modificatio...
Ivan Yakovlev, Anton Okhotnikov
Robust Sequential Hypothesis Testing with Generalized Estimating Equations
Existing sequential generalized estimating equation methodology for longitudinal and group-correlated data focuses on narrow hypotheses concerning treatment efficacy and often makes modeling assump...
Nathan T. Provost, Abdus S. Wahed
Multimodal classification of Radiation-Induced Contrast Enhancements and tumor recurrence using deep learning
The differentiation between tumor recurrence and radiation-induced contrast enhancements in post-treatment glioblastoma patients remains a major clinical challenge. Existing approaches rely on clin...
Robin Peretzke, Marlin Hanstein, Maximilian Fischer, Lars Badhi Wessel, Obada Alhalabi, Sebastian...
Chemo-dynamical reconstruction of Milky Way globular cluster progenitors: age--metallicity relations and the universality of multiple stellar populations
Globular clusters encode the hierarchical assembly history of the Milky Way and the physics of multiple stellar populations. Using homogeneous stellar parameters for 69 Galactic globular clusters d...
Carmela Lardo, David Valcin, Raul Jimenez
A Diffeomorphism Groupoid and Algebroid Framework for Discontinuous Image Registration
In this paper, we propose a novel mathematical framework for piecewise diffeomorphic image registration that involves discontinuous sliding motion using a diffeomorphism groupoid and algebroid appr...
Lili Bao, Bin Xiao, Shihui Ying, Stefan Sommer
Continuum Reverberation in Bright Quasars Using NASA/ATLAS
Variable continuum emission from AGN can be used to probe the structure of their accretion disks via reverberation mapping. Assuming a variable, hot inner light source irradiates the surrounding ac...
Zachary Steyn, Christian Wolf, Christopher Onken, Ken Smith, Ji-Jia Tang, Andjelka B. Kovacevic, ...
Origin of open clusters revealed by the evolution of the m_max$-$M_ecl relation
Using the Gaia DR3 open cluster catalog, we identified the most massive star in each observed cluster. Examining the m_max$-$M_cluster relations across different age ranges, we find that as cluster...
J. W. Zhou, Sami Dib, Pavel Kroupa
Governing Evolving Memory in LLM Agents: Risks, Mechanisms, and the Stability and Safety Governed Memory (SSGM) Framework
Long-term memory has emerged as a foundational component of autonomous Large Language Model (LLM) agents, enabling continuous adaptation, lifelong multimodal learning, and sophisticated reasoning. ...
Chingkwun Lam, Jiaxin Li, Lingfei Zhang, Kuo Zhao
Anomaly detection in time-series via inductive biases in the latent space of conditional normalizing flows
Deep generative models for anomaly detection in multivariate time-series are typically trained by maximizing data likelihood. However, likelihood in observation space measures marginal density rath...
David Baumgartner, Eliezer de Souza da Silva, Iñigo Urteaga
Mitigating the Multiplicity Burden: The Role of Calibration in Reducing Predictive Multiplicity of Classifiers
As machine learning models are increasingly deployed in high-stakes environments, ensuring both probabilistic reliability and prediction stability has become critical. This paper examines the inter...
Mustafa Cavus
Compression Favors Consistency, Not Truth: When and Why Language Models Prefer Correct Information
Why do language models sometimes prefer correct statements even when trained on mixed-quality data? We introduce the Compression--Consistency Principle: next-token prediction favors hypotheses that...
Konstantin Krestnikov
Merger-driven buildup of the $M_{\rm BH}$ - $M_*$ relation bridging high-$z$ overmassive black holes with the local relation
The origin of the mass scaling relation between supermassive black holes (SMBHs, $M_{\rm BH}$) and galaxies ($M_*$) remains a key open question. Rather than invoking AGN feedback, a non-causal mech...
Takumi S. Tanaka, John D. Silverman, Kazuhiro Shimasaku, Knud Jahnke, Junyao Li, Makoto Ando
VTEdit-Bench: A Comprehensive Benchmark for Multi-Reference Image Editing Models in Virtual Try-On
As virtual try-on (VTON) continues to advance, a growing number of real-world scenarios have emerged, pushing beyond the ability of the existing specialized VTON models. Meanwhile, universal multi-...
Xiaoye Liang, Zhiyuan Qu, Mingye Zou, Jiaxin Liu, Lai Jiang, Mai Xu, Yiheng Zhu
Inverse $t$-source problem and a strict positivity property for coupled subdiffusion systems
This article is concerned with the inverse problem on determining the temporal component of the source term in a coupled system of time-fractional diffusion equations by single point observation. U...
Mohamed BenSalah, Yikan Liu
In the LLM era, Word Sense Induction remains unsolved
In the absence of sense-annotated data, word sense induction (WSI) is a compelling alternative to word sense disambiguation, particularly in low-resource or domain-specific settings. In this paper,...
Anna Mosolova, Marie Candito, Carlos Ramisch
Compact LABFM: a framework for meshless methods with spectral-like resolving power
Meshless methods are often used in numerical simulations of systems of partial differential equations (PDEs), particularly those which involve complex geometries or free surfaces. Here we present a...
Henry M. Broadley, Steven J. Lind, Jack R. C. King