arXiv:2602.10125v1 Announce Type: new Abstract: Network segmentation is a popular security practice for limiting lateral movement, yet practitioners lack a metric to measure how segmented a network actually is. We model a network as a graph and study segmentedness as a property captured by the global edge density that can be estimated from sampled node pairs. Then, we derive an estimator and evaluate its uncertainty using confidence intervals. For a 95% confidence interval with a margin-of-error of $pm […]

Ver mais

In this work, we present a generic classification of a class of n-dimensional piecewise smooth vector fields known as refractive systems. Our purpose is to characterize the local structural stability of refractive fold-folds of elliptic type. To this end, we reduce the study of such piecewise smooth vector fields to the analysis of the first return maps and their structural stability. Normal forms for such systems are also provided.

Ver mais

The rapid evolution of Large Language Models (LLMs) from static text generators to autonomous agents has revolutionized their ability to perceive, reason, and act within complex environments. However, this transition from single-model inference to System Engineering Security introduces unique structural vulnerabilities—specifically instruction-data conflation, persistent cognitive states, and untrusted coordination—that extend beyond traditional adversarial robustness. To address the fragmented nature of the existing literature, this article presents a comprehensive and systematic survey of the security landscape for LLM-based agents. […]

Ver mais

City-scale, in-the-wild Augmented Reality (AR) learning paths must remain operable under Bring Your Own Device (BYOD) heterogeneity, outdoor tracking degradation, public-space safety, and interruption recovery. This study conceptualizes the Art Nouveau Path as an AR learning service and makes a theoretical contribution by proposing a Determinant-driven Requirements traceability model that treats implementation Determinants as Requirements signals and links them to testable Requirements, transfer Artefacts, and evidence anchors for replication. Methods combined 8 Points of Interest (POIs) and 36 […]

Ver mais

arXiv:2305.19640v3 Announce Type: replace Abstract: In this paper, we are concerned with the generalization performance of non-parametric estimation for pairwise learning. Most of the existing work requires the hypothesis space to be convex or a VC-class, and the loss to be convex. However, these restrictive assumptions limit the applicability of the results in studying many popular methods, especially kernel methods and neural networks. We significantly relax these restrictive assumptions and establish a sharp oracle inequality of the empirical […]

Ver mais

Table of Links Abstract and 1. Introduction Related Work and Background Analysis 3.1 Limitations about Existing ReLUficatio 3.2 dReLU Are Neurons in Expert still Sparsely Activated? dReLU Sparsification Experiments Results 6.1 Downstream Tasks Performance 6.2 Sparsity of Sparsified Models Practical Inference Speedup Evaluation 7.1 Experiments Setting 7.2 Pure CPU Inference and 7.3 Hybrid GPU-CPU Inference 7.4 Deploy LLMs on mobile phones Conclusion and References A. Appendix / supplemental material B. Limitation C. Broader Impact 7.1 Experiments Setting Baselines. […]

Ver mais

arXiv:2601.16397v1 Announce Type: new Abstract: Trustworthy clinical summarization requires not only fluent generation but also transparency about where each statement comes from. We propose a training-free framework for generation-time source attribution that leverages decoder attentions to directly cite supporting text spans or images, overcoming the limitations of post-hoc or retraining-based methods. We introduce two strategies for multimodal attribution: a raw image mode, which directly uses image patch attentions, and a caption-as-span mode, which substitutes images with generated captions […]

Ver mais

arXiv:2602.22296v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has improved the reasoning abilities of large language models (LLMs) on mathematics and programming tasks, but standard approaches that optimize single-attempt accuracy can inadvertently suppress response diversity across repeated attempts, narrowing exploration and overlooking underrepresented strategies. We introduce UpSkill, a training time method that adapts Mutual Information Skill Learning (MISL) to LLMs for optimizing pass@k correctness. We propose a novel reward that we implement within Group Relative […]

Ver mais