arXiv subjects

Wenbo Sun

Publications and source records attributed to Wenbo Sun.

At least 19 recordsLinked to original sources

arXiv2026

Spatial Entropy based Partitioning for Spatiotemporal Graph Unlearning

Spatiotemporal graphs underpin applications such as traffic forecasting, weather forecasting, and healthcare monitoring. Privacy regulations such as the GDPR and the CCPA require the complete removal of unauthorized data from trained models, but achieving this on a spatiotemporal graph is difficult: because information propagates globally through both spatial and temporal message passing, fully erasing a node's influence forces costly full-graph retraining. ST-graph unlearning requires both exactness and efficiency. We propose IsleNet, which uses spatial-entropy-guided partitioning to create balanced, locally coherent subgraphs and reconnects them with lightweight virtual edges. Upon an unlearning request, only the affected subgraph encoder and virtual-edge layer are retrained, ensuring exact removal with low cost. Experiments on four real-world benchmarks show that IsleNet attains up to 94% of full-graph accuracy while reducing unlearning time by up to an order of magnitude. Our code is publicly available at https://github.com/wenlu-lab/STGraphUnlearning.

cs.LG↗

arXiv2026

Unlearning on Spatio-Temporal Graphs through Subgraph Virtual Edge Reconstruction

Spatio-temporal graphs are widely used in modeling complex dynamic processes such as temporal forecasting, molecular dynamics, and healthcare monitoring. Recently, stringent privacy regulations such as GDPR and CCPA have introduced significant new challenges for existing spatio-temporal graph models, requiring complete unlearning of unauthorized data. Since each node in a spatio-temporal graph diffuses information globally across both spatial and temporal dimensions, existing unlearning methods primarily designed for static graphs and localized data removal cannot efficiently erase a single node without incurring costs nearly equivalent to full model retraining. To address this, we propose CallosumNet, a spatio-temporal graph unlearning framework biologically inspired by the corpus callosum structure. CallosumNet makes two key technical contributions: (1) it reconstructs subgraphs using biologically-inspired virtual edges; and (2) it restores interlinked spatio-temporal dependencies among subgraphs via a lightweight meta-graph integration layer. Empirical results on four diverse real-world datasets show that CallosumNet achieves complete unlearning while maintaining accuracy very close to the gold model. The code is publicly available at https://github.com/wenlu-lab/STGraphUnlearning.

cs.LG↗

arXiv2026

Adversarial Attacks on Deep OCR Systems

Deep-OCR (DeepSeek-OCR) advances document recognition by treating the visual modality as an optical compression medium, enabling long-context OCR at low token cost. However, its increased complexity may introduce new security vulnerabilities. In this paper, we present, to the best of our knowledge, the first pure black-box adversarial attack against a generative OCR vision-language model, where only the decoded string can be queried and no gradients, logits, or model internals are available. We recast the attack as a zeroth-order optimization problem driven by a bounded scalar loss defined directly on the string output via sequence similarity, and estimate the gradient with a random-direction finite-difference scheme whose query cost is independent of the image dimension. An Adam update with ell_infinity projection yields imperceptible perturbations for both untargeted and targeted objectives. Pilot experiments on Deep-OCR validate the string-only attack and evaluation pipeline and expose severe qualitative decoder failures, including repetition, truncation, and prompt leakage. They also show that controlled targeted rewriting remains substantially harder than untargeted degradation; we avoid claiming targeted success until the pre-registered evaluation is complete.

cs.CR↗

arXiv2026

Structure of 2-step nilpotent ergodic averages for distinct-degree polynomials

We investigate manifestations of the Nilpotent Heuristic, which posits that recurrence and convergence phenomena known for measure-preserving $\mathbb{Z}^D$-systems extend to nilpotent group actions. Our main results establish seminorm estimates and limiting formulas for multiple ergodic averages arising from actions of 2-step nilpotent groups. In particular, if $T_1,\ldots,T_\ell$ are totally ergodic and generate a 2-step nilpotent group, then \[ \lim_{N\to\infty}\frac{1}{N}\sum_{n=1}^N T_1^n f_1 \cdots T_\ell^{n^\ell}f_\ell = \prod_{j=1}^{\ell}\int f_j\,d\mu \] in the $L^{2}$ norm for all bounded functions $f_{1},\dots,f_{\ell}$; the same holds for any distinct-degree polynomial iterates. We also obtain popular-common-difference versions of the polynomial Szeme\'edi theorem in the same setting. In a different direction, our approach allows us to completely resolve the joint ergodicity conjecture for multidimensional polynomials and $\mathbb Z^D$-systems; we also present an example showing that, surprisingly enough, the 2-step nilpotent analog fails. We conclude with many open problems concerning joint ergodicity, seminorm estimates, and the structure theory of nilpotent systems.

math.DS↗

arXiv2026

Superconductivity in Al-based high-entropy alloys TiHfNbTaAl and TaNbHfZrAl

Since the first report of a high entropy alloy (HEA) superconductor in 2014, HEAs have continued to captivate the interest of superconducting researchers. Owing to the significant degree of disorder inherent in these systems, they serve as exemplary models for examining the properties of materials that exist in states intermediate between crystalline and amorphous structures. Here we present the superconductivity properties and crystal structure of TaNbHfZrAl and TiHfNbTaAl HEAs, which both have the VEC of 4.2 and body centered cubic (BCC) structure. Through resistivity, magnetic, and specific heat measurements, we prove that both samples are the bulk type-II superconductors with a critical temperature Tc of 5.5 K for TaNbHfZrAl and Tc of 3.2 K for TiHfNbTaAl. The Tc of HEA superconductors is influenced by the VEC and the element composition. And the incorporation of Al in high disorder HEA superconductors causes a more crystallinelike Tc dependence.

cond-mat.supr-con↗

arXiv2026

Beyond the Purcell Effect: Controlling Pure Quantum Dephasing with Spin Noise Metasurfaces

One central theme in quantum photonics is tailoring the interactions between atoms/spins and their electromagnetic (EM) environments. Considerable effort has focused on engineering spontaneous emission by shaping EM environments, known as the Purcell effect. However, photonic environment control of pure dephasing, which is a complementary paradigm of non-unitary atom/spin couplings with EM environments, remains largely unexplored. Here, we introduce a nanophotonic approach to modify qubit pure dephasing dynamics. Unlike Purcell engineering that tailors photonic environments at qubit resonance frequencies (typically optical/near-infrared), we develop ultra-subwavelength spin noise metasurfaces for efficient broadband control of low-frequency (e.g., $\sim$MHz) photonic environments far off-resonant with atoms/spins for dephasing engineering. We experimentally demonstrate our approach using lithographically defined CoFeB metasurfaces and shallow nitrogen-vacancy (NV) centers in diamond. Instead of modified spontaneous emission, we observe modified NV pure dephasing dynamics near different spin noise metasurfaces. We further isolate metasurface-controlled dephasing from other dephasing mechanisms (e.g., spin bath) by measuring the NV ensemble dephasing noise spectrum with dynamical decoupling spectral decomposition techniques. Our results establish a new frontier in engineering quantum light-matter interactions with nanophotonic structures.

quant-ph↗

arXiv2026

Correlated Quantum Dephasometry: Symmetry-Resolved Noise Spectroscopy of Two-Dimensional Superconductors and Altermagnets

Symmetry-resolved spectroscopies, such as angle-resolved photoemission spectroscopy and polarization-resolved Raman, are central for quantum material characterization, yet remain challenging at nanoscale dimensions and low frequencies. Here, we propose correlated quantum dephasometry, which enables symmetry resolved quantum noise spectroscopy of materials at nanoscale and low ($\sim$MHz) frequencies via correlated dephasing of two spin qubits near materials. Our approach leverages the finite-range spatial structures of nonlocal near-field noise correlations to isolate rotational symmetry of the material response in momentum space beyond single qubit capabilities. We apply our approach to two-dimensional (2D) superconductors, and predict clear fingerprints that discriminate s-, d-, and g-wave symmetry of the superconducting gap. To highlight the generality, we further show that the same framework resolves 2D s-wave antiferromagnets and d-wave altermagnets. Our results establish correlated quantum dephasometry as a nanoscale, low-frequency complement for symmetry-resolved spectroscopy applicable to a broad class of quantum materials.

quant-ph↗

arXiv2026

Partition regularity in imaginary quadratic rings of integers

We obtain partition regularity results for homogeneous quadratic equations whose parametrized solutions admit nice factorizations into linear forms over rings of integers of imaginary quadratic fields. To do so, we develop number-theoretic results of independent interest on such fields, such as a characterization for aperiodic completely multiplicative functions, the Turán-Kubilius inequality, and a new concentration estimate for multiplicative functions.

math.CO↗

arXiv2026

Probing near-field EM fluctuations in superparamagnetic CoFeB with NV quantum dephasometry

Superparamagnetism in nanoscale magnetic layers is a critical property for a wide range of spintronic-based sensor and computing applications. While conventional magnetization measurements can detect superparamagnetic signatures, they often require the application of high perturbative fields and are difficult to implement for magnetic layers integrated within functional devices. In this study, we non-invasively investigate the superparamagnetic spin dynamics of a nanoscale CoFeB layer of thickness 1.1 nm, deposited on a diamond substrate, by probing its low-frequency near-field electromagnetic (EM) fluctuations using nitrogen-vacancy (NV) centers-based quantum dephasometry. Our measurements reveal an unconventional, non-monotonic temperature dependence of the dephasing time of NV centers, which we attribute to EM fluctuations produced by thermally driven superparamagnetic domain flipping in CoFeB. Our findings are further supported by the theoretical interpretation of the dephasing dynamics of NV centers and the complementary SQUID-based magnetization characterizations of the CoFeB layer. Additionally, exploiting the technique of NV dephasometry, we extract the spectral density of the EM fluctuations in CoFeB, which is used to isolate different components of the EM fluctuations acting on NV centers. We also measure the CoFeB-to-NV distance-dependent coherence times of NV centers to investigate the effect of the dimensionality of the CoFeB layer on the generated near-field EM fluctuations. These results provide critical insight into the magnetization dynamics and near-field EM environment of nanoscale magnetic layers. It also has significant implications for the development of hybrid quantum spintronic devices and applications involving nanoscale opto-magnetic materials.

cond-mat.mes-hall↗

arXiv2025

Seminorm estimates and joint ergodicity for pairwise independent Hardy sequences

We develop a robust structure theory for multiple ergodic averages of commuting transformations along Hardy sequences of polynomial growth. We then apply it to derive a number of novel results on joint ergodicity, recurrence and convergence. In particular, we prove joint ergodicity for (a) pairwise independent Hardy sequences and weakly mixing transformations, (b) strongly independent Hardy sequences and ergodic transformations, (c) strongly irrationally independent Hardy sequences and totally ergodic transformations. We use these joint ergodicity results to provide new recurrence results for multidimensional patterns along strongly independent Hardy sequences, showing for instance that all subsets of $\mathbb{Z}^2$ of positive upper density contain patterns of the form $$ (m_1, m_2),\; (m_1 + \lfloor n^{\sqrt{2}}\rfloor, m_2),\; (m_1, m_2 + \lfloor n^{\sqrt{2}} + n^{1/2}\rfloor).$$ Last but not least, we positively resolve the joint ergodicity classification problem for pairwise independent Hardy sequences, of which the aforementioned families are special cases. While building on recent technical advances (e.g. PET coefficient tracking schemes and joint ergodicity criteria), our work introduces a number of technical developments of its own. We construct a suitable generalization of Host-Kra and box seminorms that quantitatively control ergodic averages along Hardy sequences. We subsequently use them to obtain Host-Kra seminorm estimates for averages along all pairwise independent Hardy sequences. Furthermore, we develop an ergodic version of the quantitative concatenation argument that has recently found extensive use in combinatorics, number theory and harmonic analysis. Lastly, we obtain new simultaneous Taylor approximations for Hardy sequences, a crucial ingredient to deal with the aforementioned classes of Hardy sequences.

math.DS↗

arXiv2025

AquaSentinel: Next-Generation AI System Integrating Sensor Networks for Urban Underground Water Pipeline Anomaly Detection via Collaborative MoE-LLM Agent Architecture

Underground pipeline leaks and infiltrations pose significant threats to water security and environmental safety. Traditional manual inspection methods provide limited coverage and delayed response, often missing critical anomalies. This paper proposes AquaSentinel, a novel physics-informed AI system for real-time anomaly detection in urban underground water pipeline networks. We introduce four key innovations: (1) strategic sparse sensor deployment at high-centrality nodes combined with physics-based state augmentation to achieve network-wide observability from minimal infrastructure; (2) the RTCA (Real-Time Cumulative Anomaly) detection algorithm, which employs dual-threshold monitoring with adaptive statistics to distinguish transient fluctuations from genuine anomalies; (3) a Mixture of Experts (MoE) ensemble of spatiotemporal graph neural networks that provides robust predictions by dynamically weighting model contributions; (4) causal flow-based leak localization that traces anomalies upstream to identify source nodes and affected pipe segments. Our system strategically deploys sensors at critical network junctions and leverages physics-based modeling to propagate measurements to unmonitored nodes, creating virtual sensors that enhance data availability across the entire network. Experimental evaluation using 110 leak scenarios demonstrates that AquaSentinel achieves 100% detection accuracy. This work advances pipeline monitoring by demonstrating that physics-informed sparse sensing can match the performance of dense deployments at a fraction of the cost, providing a practical solution for aging urban infrastructure.

cs.CE↗

arXiv2025

Spatio-Temporal Graph Unlearning

Spatio-temporal graphs are widely used in modeling complex dynamic processes such as traffic forecasting, molecular dynamics, and healthcare monitoring. Recently, stringent privacy regulations such as GDPR and CCPA have introduced significant new challenges for existing spatio-temporal graph models, requiring complete unlearning of unauthorized data. Since each node in a spatio-temporal graph diffuses information globally across both spatial and temporal dimensions, existing unlearning methods primarily designed for static graphs and localized data removal cannot efficiently erase a single node without incurring costs nearly equivalent to full model retraining. Therefore, an effective approach for complete spatio-temporal graph unlearning is a pressing need. To address this, we propose CallosumNet, a divide-and-conquer spatio-temporal graph unlearning framework inspired by the corpus callosum structure that facilitates communication between the brain's two hemispheres. CallosumNet incorporates two novel techniques: (1) Enhanced Subgraph Construction (ESC), which adaptively constructs multiple localized subgraphs based on several factors, including biologically-inspired virtual ganglions; and (2) Global Ganglion Bridging (GGB), which reconstructs global spatio-temporal dependencies from these localized subgraphs, effectively restoring the full graph representation. Empirical results on four diverse real-world datasets show that CallosumNet achieves complete unlearning with only 1%-2% relative MAE loss compared to the gold model, significantly outperforming state-of-the-art baselines. Ablation studies verify the effectiveness of both proposed techniques.

cs.LG↗

arXiv2025

Quality-Ensured In-Situ Process Monitoring with Deep Canonical Correlation Analysis

This paper proposes a deep learning-based approach for in-situ process monitoring that captures nonlinear relationships between in-control high-dimensional process signature signals and offline product quality data. Specifically, we introduce a Deep Canonical Correlation Analysis (DCCA)-based framework that enables the joint feature extraction and correlation analysis of multi-modal data sources, such as optical emission spectra and CT scan images, which are collected in advanced manufacturing processes. This unified framework facilitates online quality monitoring by learning quality-oriented representations without requiring labeled defective samples and avoids the non-normality issues that often degrade traditional control chart-based monitoring techniques. We provide theoretical guarantees for the method's stability and convergence and validate its effectiveness and practical applicability through simulation experiments and a real-world case study on Direct Metal Deposition (DMD) additive manufacturing.

stat.AP↗

arXiv2025

CEBench: A Benchmarking Toolkit for the Cost-Effectiveness of LLM Pipelines

Online Large Language Model (LLM) services such as ChatGPT and Claude 3 have transformed business operations and academic research by effortlessly enabling new opportunities. However, due to data-sharing restrictions, sectors such as healthcare and finance prefer to deploy local LLM applications using costly hardware resources. This scenario requires a balance between the effectiveness advantages of LLMs and significant financial burdens. Additionally, the rapid evolution of models increases the frequency and redundancy of benchmarking efforts. Existing benchmarking toolkits, which typically focus on effectiveness, often overlook economic considerations, making their findings less applicable to practical scenarios. To address these challenges, we introduce CEBench, an open-source toolkit specifically designed for multi-objective benchmarking that focuses on the critical trade-offs between expenditure and effectiveness required for LLM deployments. CEBench allows for easy modifications through configuration files, enabling stakeholders to effectively assess and optimize these trade-offs. This strategic capability supports crucial decision-making processes aimed at maximizing effectiveness while minimizing cost impacts. By streamlining the evaluation process and emphasizing cost-effectiveness, CEBench seeks to facilitate the development of economically viable AI solutions across various industries and research fields. The code and demonstration are available in https://github.com/amademicnoboday12/CEBench.

cs.PF↗

arXiv2025

TranSQL+: Serving Large Language Models with SQL on Low-Resource Hardware

Deploying Large Language Models (LLMs) on resource-constrained devices remains challenging due to limited memory, lack of GPUs, and the complexity of existing runtimes. In this paper, we introduce TranSQL+, a template-based code generator that translates LLM computation graphs into pure SQL queries for execution in relational databases. Without relying on external libraries, TranSQL+, leverages mature database features, such as vectorized execution and out-of-core processing, for efficient inference. We further propose a row-to-column (ROW2COL) optimization that improves join efficiency in matrix operations. Evaluated on Llama3-8B and DeepSeekMoE models, TranSQL+ achieves up to 20x lower prefill latency and 4x higher decoding speed compared to DeepSpeed Inference and Llama.cpp in low-memory and CPU-only configurations. Our results highlight relational databases as a practical environment for LLMs on low-resource hardware.

cs.DB↗

arXiv2025

SouLLMate: An Adaptive LLM-Driven System for Advanced Mental Health Support and Assessment, Based on a Systematic Application Survey

Mental health issues significantly impact individuals' daily lives, yet many do not receive the help they need even with available online resources. This study aims to provide accessible, stigma-free, personalized, and real-time mental health support through cutting-edge AI technologies. It makes the following contributions: (1) Conducting an extensive survey of recent mental health support methods to identify prevalent functionalities and unmet needs. (2) Introducing SouLLMate, an adaptive LLM-driven system that integrates LLM technologies, Chain, Retrieval-Augmented Generation (RAG), prompt engineering, and domain knowledge. This system offers advanced features such as Suicide Risk Detection and Proactive Guidance Dialogue, and utilizes RAG for personalized profile uploads and Conversational Information Extraction. (3) Developing novel evaluation approaches to assess preliminary assessments and suicide risk detection, utilizing annotated real-life interview data and professionally labeled datasets indicating suicide tendencies. (4) Proposing Key Indicator Summarization (KIS) and Proactive Questioning Strategy (PQS) methods to enhance model performance and usability through context-sensitive response adjustments and semantic coherence evaluations. This study contributes to advancing mental health support technologies, potentially improving the accessibility and effectiveness of mental health care globally.

cs.HC↗

arXiv2025

SouLLMate: An Application Enhancing Diverse Mental Health Support with Adaptive LLMs, Prompt Engineering, and RAG Techniques

Mental health issues significantly impact individuals' daily lives, yet many do not receive the help they need even with available online resources. This study aims to provide diverse, accessible, stigma-free, personalized, and real-time mental health support through cutting-edge AI technologies. It makes the following contributions: (1) Conducting an extensive survey of recent mental health support methods to identify prevalent functionalities and unmet needs. (2) Introducing SouLLMate, an adaptive LLM-driven system that integrates LLM technologies, Chain, Retrieval-Augmented Generation (RAG), prompt engineering, and domain knowledge. This system offers advanced features such as Risk Detection and Proactive Guidance Dialogue, and utilizes RAG for personalized profile uploads and Conversational Information Extraction. (3) Developing novel evaluation approaches for preliminary assessments and risk detection via professionally annotated interview data and real-life suicide tendency data. (4) Proposing the Key Indicator Summarization (KIS), Proactive Questioning Strategy (PQS), and Stacked Multi-Model Reasoning (SMMR) methods to enhance model performance and usability through context-sensitive response adjustments, semantic coherence evaluations, and enhanced accuracy of long-context reasoning in language models. This study contributes to advancing mental health support technologies, potentially improving the accessibility and effectiveness of mental health care globally.

cs.CL↗

arXiv2025

A Layered Multi-Expert Framework for Long-Context Mental Health Assessments

Long-form mental health assessments pose unique challenges for large language models (LLMs), which often exhibit hallucinations or inconsistent reasoning when handling extended, domain-specific contexts. We introduce Stacked Multi-Model Reasoning (SMMR), a layered framework that leverages multiple LLMs and specialized smaller models as coequal 'experts'. Early layers isolate short, discrete subtasks, while later layers integrate and refine these partial outputs through more advanced long-context models. We evaluate SMMR on the DAIC-WOZ depression-screening dataset and 48 curated case studies with psychiatric diagnoses, demonstrating consistent improvements over single-model baselines in terms of accuracy, F1-score, and PHQ-8 error reduction. By harnessing diverse 'second opinions', SMMR mitigates hallucinations, captures subtle clinical nuances, and enhances reliability in high-stakes mental health assessments. Our findings underscore the value of multi-expert frameworks for more trustworthy AI-driven screening.

cs.CL↗