arXiv subjects

Wei Lu

Publications and source records attributed to Wei Lu.

At least 163 records · Page 9Linked to original sources

arXiv2025

OptScaler: A Collaborative Framework for Robust Autoscaling in the Cloud

Autoscaling is a critical mechanism in cloud computing, enabling the autonomous adjustment of computing resources in response to dynamic workloads. This is particularly valuable for co-located, long-running applications with diverse workload patterns. The primary objective of autoscaling is to regulate resource utilization at a desired level, effectively balancing the need for resource optimization with the fulfillment of Service Level Objectives (SLOs). Many existing proactive autoscaling frameworks may encounter prediction deviations arising from the frequent fluctuations of cloud workloads. Reactive frameworks, on the other hand, rely on realtime system feedback, but their hysteretic nature could lead to violations of stringent SLOs. Hybrid frameworks, while prevalent, often feature independently functioning proactive and reactive modules, potentially leading to incompatibility and undermining the overall decision-making efficacy. In addressing these challenges, we propose OptScaler, a collaborative autoscaling framework that integrates proactive and reactive modules through an optimization module. The proactive module delivers reliable future workload predictions to the optimization module, while the reactive module offers a self-tuning estimator for real-time updates. By embedding a Model Predictive Control (MPC) mechanism and chance constraints into the optimization module, we further enhance its robustness. Numerical results have demonstrated the superiority of our workload prediction model and the collaborative framework, leading to over a 36% reduction in SLO violations compared to prevalent reactive, proactive, or hybrid autoscalers. Notably, OptScaler has been successfully deployed at Alipay, providing autoscaling support for the world-leading payment platform.

math.OC↗

arXiv2025

TxnSails: Achieving Serializable Transaction Scheduling with Self-Adaptive Isolation Level Selection

Achieving the serializable isolation level, regarded as the gold standard for transaction processing, is costly. Recent studies reveal that adjusting specific query patterns within a workload can still achieve serializability even at lower isolation levels. Nevertheless, these studies typically overlook the trade-off between the performance advantages of lower isolation levels and the overhead required to maintain serializability, potentially leading to suboptimal isolation level choices that fail to maximize performance. In this paper, we present TxnSails, a middle-tier solution designed to achieve serializable scheduling with self-adaptive isolation level selection. First, TxnSails incorporates a unified concurrency control algorithm that achieves serializability at lower isolation levels with minimal additional overhead. Second, TxnSails employs a deep learning method to characterize the trade-off between the performance benefits and overhead associated with lower isolation levels, thus predicting the optimal isolation level. Finally, TxnSails implements a cross-isolation validation mechanism to ensure serializability during real-time isolation level transitions. Extensive experiments demonstrate that TxnSails outperforms state-of-the-art solutions by up to 26.7x and PostgreSQL's serializable isolation level by up to 4.8x.

cs.DB↗

arXiv2025

Surveying Attitudinal Alignment Between Large Language Models Vs. Humans Towards 17 Sustainable Development Goals

Large Language Models (LLMs) have emerged as potent tools for advancing the United Nations' Sustainable Development Goals (SDGs). However, the attitudinal disparities between LLMs and humans towards these goals can pose significant challenges. This study conducts a comprehensive review and analysis of the existing literature on the attitudes of LLMs towards the 17 SDGs, emphasizing the comparison between their attitudes and support for each goal and those of humans. We examine the potential disparities, primarily focusing on aspects such as understanding and emotions, cultural and regional differences, task objective variations, and factors considered in the decision-making process. These disparities arise from the underrepresentation and imbalance in LLM training data, historical biases, quality issues, lack of contextual understanding, and skewed ethical values reflected. The study also investigates the risks and harms that may arise from neglecting the attitudes of LLMs towards the SDGs, including the exacerbation of social inequalities, racial discrimination, environmental destruction, and resource wastage. To address these challenges, we propose strategies and recommendations to guide and regulate the application of LLMs, ensuring their alignment with the principles and goals of the SDGs, and therefore creating a more just, inclusive, and sustainable future.

cs.CY↗

arXiv2025

Boundary-enhanced time series data imputation with long-term dependency diffusion models

Data imputation is crucial for addressing challenges posed by missing values in multivariate time series data across various fields, such as healthcare, traffic, and economics, and has garnered significant attention. Among various methods, diffusion model-based approaches show notable performance improvements. However, existing methods often cause disharmonious boundaries between missing and known regions and overlook long-range dependencies in missing data estimation, leading to suboptimal results. To address these issues, we propose a Diffusion-based time Series Data Imputation (DSDI) framework. We develop a weight-reducing injection strategy that incorporates the predicted values of missing points with reducing weights into the reverse diffusion process to mitigate boundary inconsistencies. Further, we introduce a multi-scale S4-based U-Net, which combines hierarchical information from different levels via multi-resolution integration to capture long-term dependencies. Experimental results demonstrate that our model outperforms existing imputation methods.

cs.LG↗

arXiv2025

Local Detection of Enhanced Hot Electron Scattering in InSb/CdTe Heterostructure Interface

The InSb/CdTe heterojunction structure, characterized by low effective mass and high electron mobility, exhibits interfacial energy band bending, leading to the Rashba spin-orbit coupling effect and nonreciprocal transport, which makes its suitable for spintronic devices with broad applications in logic and storage fields. However, the complex heterojunction interfaces of InSb/CdTe, composed of group III-V and group II-VI semiconductors, are prone to interdiffusion. Therefore, characterization and study of the interfacial properties of InSb/CdTe heterojunctions are crucial for the growth improvement of the InSb/CdTe material system as well as its application in the field of spintronics. In this study, a novel scanning probe microscope, called a scanning noise microscope, was applied to visualize hot electron scattering in InSb/CdTe nano-devices. The results demonstrated that the near-field signal originates from the Coulomb scattering of charged ions on electrons at the interface of the embedded layer heterojunction. This real-space, nondestructive characterization of the heterojunction interface properties offers a new tool for enhancing the performance of heterojunctions.

physics.app-ph↗

arXiv2025

Vulnerabilities Analysis and Secure Controlling for Unmanned Aerial System Based on Reactive Synthesis

Complex Cyber-Physical System (CPS) such as Unmanned Aerial System (UAS) got rapid development these years, but also became vulnerable to GPS spoofing, packets injection, buffer-overflow and other malicious attacks. Ensuring the behaviors of UAS always keeping secure no matter how the environment changes, would be a prospective direction for UAS security. This paper aims at introducing a pattern-based framework to describe the security properties of UAS, and presenting a reactive synthesis-based approach to implement the automatic generation of secure UAS controller. First, we study the operating mechanism of UAS and construct a high-level model consisting of actuator and monitor. Besides, we analyze the security threats of UAS from the perspective of hardware, software and cyber physics, and then summarize the corresponding specification patterns of security properties with LTL formulas. With the UAS model and security specification patterns, automatons for controller can be constructed by General Reactivity of Rank 1 (GR(1)) synthesis algorithm, which is a two-player game process between Unmanned Aerial Vehicle (UAV) and its environment. Finally, we experimented under the Ardupilot simulation platform to test the effectiveness of our method.

cs.FL↗

arXiv2024

Applying Graph Explanation to Operator Fusion

Layer fusion techniques are critical to improving the inference efficiency of deep neural networks (DNN) for deployment. Fusion aims to lower inference costs by reducing data transactions between an accelerator's on-chip buffer and DRAM. This is accomplished by grouped execution of multiple operations like convolution and activations together into single execution units - fusion groups. However, on-chip buffer capacity limits fusion group size and optimizing fusion on whole DNNs requires partitioning into multiple fusion groups. Finding the optimal groups is a complex problem where the presence of invalid solutions hampers traditional search algorithms and demands robust approaches. In this paper we incorporate Explainable AI, specifically Graph Explanation Techniques (GET), into layer fusion. Given an invalid fusion group, we identify the operations most responsible for group invalidity, then use this knowledge to recursively split the original fusion group via a greedy tree-based algorithm to minimize DRAM access. We pair our scheme with common algorithms and optimize DNNs on two types of layer fusion: Line-Buffer Depth First (LBDF) and Branch Requirement Reduction (BRR). Experiments demonstrate the efficacy of our scheme on several popular and classical convolutional neural networks like ResNets and MobileNets. Our scheme achieves over 20% DRAM Access reduction on EfficientNet-B3.

cs.LG↗

arXiv2024

Clifford algebra Cl(0,6) approach to beyond the standard model and naturalness problems

Is there more to Dirac's gamma matrices than meets the eye? It turns out that gamma zero can be factorized into a product of three operators. This revelation facilitates the expansion of Dirac's space-time algebra to Clifford algebra Cl(0,6). The resultant rich geometric structure can be leveraged to establish a combined framework of the standard model and gravity, wherein a gravi-weak interaction between the extended vierbein field and the extended weak gauge field is allowed. In conjunction with the composite Higgs model, we examine the vierbein field as a Cooper-pair-like fermion-antifermion condensation. Quantum gravity is realized indirectly via the quantized standard model spinor fields which underlie the composite space-time metric. We propose that the fundamental energy scales of the universe including the Planck scale are emergent and resulted from quantum condensations, thus possibly addressing the cosmological constant problem through an unconventional multi-scale renormalization procedure for multiplications of divergent Feynman integrals. The Clifford algebra approach also permits a weaker form of charge conjugation without particle-antiparticle interchange, leading to a Majorana-type mass that conserves lepton number. Additionally, with reshuffling the traditional quark-lepton pairing pattern of three generations of fermions, we explore a three-Higgs-doublet model with Higgs VEVs 246 GeV, 42 GeV and 2.5 GeV which could explain the mass hierarchies of fermions.

physics.gen-ph↗

arXiv2024

SUMI-IFL: An Information-Theoretic Framework for Image Forgery Localization with Sufficiency and Minimality Constraints

Image forgery localization (IFL) is a crucial technique for preventing tampered image misuse and protecting social safety. However, due to the rapid development of image tampering technologies, extracting more comprehensive and accurate forgery clues remains an urgent challenge. To address these challenges, we introduce a novel information-theoretic IFL framework named SUMI-IFL that imposes sufficiency-view and minimality-view constraints on forgery feature representation. First, grounded in the theoretical analysis of mutual information, the sufficiency-view constraint is enforced on the feature extraction network to ensure that the latent forgery feature contains comprehensive forgery clues. Considering that forgery clues obtained from a single aspect alone may be incomplete, we construct the latent forgery feature by integrating several individual forgery features from multiple perspectives. Second, based on the information bottleneck, the minimality-view constraint is imposed on the feature reasoning network to achieve an accurate and concise forgery feature representation that counters the interference of task-unrelated features. Extensive experiments show the superior performance of SUMI-IFL to existing state-of-the-art methods, not only on in-dataset comparisons but also on cross-dataset comparisons.

cs.CV↗

arXiv2024

Spectral-Temporal Fusion Representation for Person-in-Bed Detection

This study is based on the ICASSP 2025 Signal Processing Grand Challenge's Accelerometer-Based Person-in-Bed Detection Challenge, which aims to determine bed occupancy using accelerometer signals. The task is divided into two tracks: "in bed" and "not in bed" segmented detection, and streaming detection, facing challenges such as individual differences, posture variations, and external disturbances. We propose a spectral-temporal fusion-based feature representation method with mixup data augmentation, and adopt Intersection over Union (IoU) loss to optimize detection accuracy. In the two tracks, our method achieved outstanding results of 100.00% and 95.55% in detection scores, securing first place and third place, respectively.

eess.SP↗

arXiv2024

RaCMC: Residual-Aware Compensation Network with Multi-Granularity Constraints for Fake News Detection

Multimodal fake news detection aims to automatically identify real or fake news, thereby mitigating the adverse effects caused by such misinformation. Although prevailing approaches have demonstrated their effectiveness, challenges persist in cross-modal feature fusion and refinement for classification. To address this, we present a residual-aware compensation network with multi-granularity constraints (RaCMC) for fake news detection, that aims to sufficiently interact and fuse cross-modal features while amplifying the differences between real and fake news. First, a multiscale residual-aware compensation module is designed to interact and fuse features at different scales, and ensure both the consistency and exclusivity of feature interaction, thus acquiring high-quality features. Second, a multi-granularity constraints module is implemented to limit the distribution of both the news overall and the image-text pairs within the news, thus amplifying the differences between real and fake news at the news and feature levels. Finally, a dominant feature fusion reasoning module is developed to comprehensively evaluate news authenticity from the perspectives of both consistency and inconsistency. Experiments on three public datasets, including Weibo17, Politifact and GossipCop, reveal the superiority of the proposed method.

cs.CV↗

arXiv2024

Interweaving Memories of a Siamese Large Language Model

Parameter-efficient fine-tuning (PEFT) methods optimize large language models (LLMs) by modifying or introducing a small number of parameters to enhance alignment with downstream tasks. However, they can result in catastrophic forgetting, where LLMs prioritize new knowledge at the expense of comprehensive world knowledge. A promising approach to mitigate this issue is to recall prior memories based on the original knowledge. To this end, we propose a model-agnostic PEFT framework, IMSM, which Interweaves Memories of a Siamese Large Language Model. Specifically, our siamese LLM is equipped with an existing PEFT method. Given an incoming query, it generates two distinct memories based on the pre-trained and fine-tuned parameters. IMSM then incorporates an interweaving mechanism that regulates the contributions of both original and enhanced memories when generating the next token. This framework is theoretically applicable to all open-source LLMs and existing PEFT methods. We conduct extensive experiments across various benchmark datasets, evaluating the performance of popular open-source LLMs using the proposed IMSM, in comparison to both classical and leading PEFT methods. Our findings indicate that IMSM maintains comparable time and space efficiency to backbone PEFT methods while significantly improving performance and effectively mitigating catastrophic forgetting.

cs.CL↗

arXiv2024

Joint Knowledge Editing for Information Enrichment and Probability Promotion

Knowledge stored in large language models requires timely updates to reflect the dynamic nature of real-world information. To update the knowledge, most knowledge editing methods focus on the low layers, since recent probes into the knowledge recall process reveal that the answer information is enriched in low layers. However, these probes only and could only reveal critical recall stages for the original answers, while the goal of editing is to rectify model's prediction for the target answers. This inconsistency indicates that both the probe approaches and the associated editing methods are deficient. To mitigate the inconsistency and identify critical editing regions, we propose a contrast-based probe approach, and locate two crucial stages where the model behavior diverges between the original and target answers: Information Enrichment in low layers and Probability Promotion in high layers. Building upon the insights, we develop the Joint knowledge Editing for information Enrichment and probability Promotion (JEEP) method, which jointly edits both the low and high layers to modify the two critical recall stages. Considering the mutual interference and growing forgetting due to dual modifications, JEEP is designed to ensure that updates to distinct regions share the same objectives and are complementary. We rigorously evaluate JEEP by editing up to thousands of facts on various models, i.e., GPT-J (6B) and LLaMA (7B), and addressing diverse editing objectives, i.e., adding factual and counterfactual knowledge. In all tested scenarios, JEEP achieves best performances, validating the effectiveness of the revealings of our probe approach and the designs of our editing method. Our code and data are available at https://github.com/Eric8932/JEEP.

cs.CL↗

arXiv2024

Qua$^2$SeDiMo: Quantifiable Quantization Sensitivity of Diffusion Models

Diffusion Models (DM) have democratized AI image generation through an iterative denoising process. Quantization is a major technique to alleviate the inference cost and reduce the size of DM denoiser networks. However, as denoisers evolve from variants of convolutional U-Nets toward newer Transformer architectures, it is of growing importance to understand the quantization sensitivity of different weight layers, operations and architecture types to performance. In this work, we address this challenge with Qua$^2$SeDiMo, a mixed-precision Post-Training Quantization framework that generates explainable insights on the cost-effectiveness of various model weight quantization methods for different denoiser operation types and block structures. We leverage these insights to make high-quality mixed-precision quantization decisions for a myriad of diffusion models ranging from foundational U-Nets to state-of-the-art Transformers. As a result, Qua$^2$SeDiMo can construct 3.4-bit, 3.9-bit, 3.65-bit and 3.7-bit weight quantization on PixArt-$α$, PixArt-$Σ$, Hunyuan-DiT and SDXL, respectively. We further pair our weight-quantization configurations with 6-bit activation quantization and outperform existing approaches in terms of quantitative metrics and generative image quality.

cs.CV↗

arXiv2024

Time-Varying Strong Coupling and Its Induced Time Diffraction of Magnon Modes

Time-varying media break the temporal translation symmetry of wave propagation in materials, enabling advanced wave manipulations. However, this novel phenomenon has been rarely explored in magnonic systems due to the significant challenge of achieving a sudden and prominent change in magnon dispersion within materials. Here, we construct a time-varying strong coupling between two magnon modes, and observe a change in the beats of Rabi-like oscillations near the pulse edges. Using a frequency-comb spectroscopy technique developed in this work, we characterize the frequency conversion of magnon modes induced by the time-varying strong-coupling effect. Moreover, we construct time slits with adjacent time interfaces and demonstrate, for the first time, the double-slit time diffraction of magnon modes, analogous to the well-known Young's double-slit experiment. These findings rely solely on the time-varying strong magnon coupling, independent of device reconfiguration. Our results open avenues for applications such as all-magnetic mixers or on-chip GHz sources.

cond-mat.mes-hall↗

arXiv2024

PersonaMark: Personalized LLM watermarking for model protection and user attribution

The rapid advancement of customized Large Language Models (LLMs) offers considerable convenience. However, it also intensifies concerns regarding the protection of copyright/confidential information. With the extensive adoption of private LLMs, safeguarding model copyright and ensuring data privacy have become critical. Text watermarking has emerged as a viable solution for detecting AI-generated content and protecting models. However, existing methods fall short in providing individualized watermarks for each user, a critical feature for enhancing accountability and traceability. In this paper, we introduce PersonaMark, a novel personalized text watermarking scheme designed to protect LLMs' copyrights and bolster accountability. PersonaMark leverages sentence structure as a subtle carrier of watermark information and optimizes the generation process to maintain the natural output of the model. By employing a personalized hashing function, unique watermarks are embedded for each user, enabling high-quality text generation without compromising the model's performance. This approach is both time-efficient and scalable, capable of handling large numbers of users through a multi-user hashing mechanism. To the best of our knowledge, this is a pioneer study to explore personalized watermarking in LLMs. We conduct extensive evaluations across four LLMs, analyzing various metrics such as perplexity, sentiment, alignment, and readability. The results validate that PersonaMark preserves text quality, ensures unbiased watermark insertion, and offers robust watermark detection capabilities, all while maintaining the model's behavior with minimal disruption.

cs.CR↗

arXiv2024

GeoTP: Latency-aware Geo-Distributed Transaction Processing in Database Middlewares (Extended Version)

The widespread adoption of database middleware for supporting distributed transaction processing is prevalent in numerous applications, with heterogeneous data sources deployed across national and international boundaries. However, transaction processing performance significantly drops due to the high network latency between the middleware and data sources and the long lock contention span, where transactions may be blocked while waiting for the locks held by concurrent transactions. In this paper, we propose GeoTP, a latency-aware geo-distributed transaction processing approach in database middlewares. GeoTP incorporates three key techniques to enhance geo-distributed transaction performance. First, we propose a decentralized prepare mechanism, which diminishes the requirement of network round trips for distributed transactions. Second, we design a latency-aware scheduler to minimize the lock contention span by strategically postponing the lock acquisition time point. Third, heuristic optimizations are proposed for the scheduler to reduce the lock contention span further. We implemented GeoTP on Apache Shardingsphere, a state-of-the-art middleware, and extended it into Apache ScalarDB. Experimental results on YCSB and TPC-C demonstrate that GeoTP achieves up to 17.7x performance improvement over Shardingsphere.

cs.DB↗

arXiv2024

Backdoor Attacks and Countermeasures in Natural Language Processing Models: A Comprehensive Security Review

Language Models (LMs) are becoming increasingly popular in real-world applications. Outsourcing model training and data hosting to third-party platforms has become a standard method for reducing costs. In such a situation, the attacker can manipulate the training process or data to inject a backdoor into models. Backdoor attacks are a serious threat where malicious behavior is activated when triggers are present, otherwise, the model operates normally. However, there is still no systematic and comprehensive review of LMs from the attacker's capabilities and purposes on different backdoor attack surfaces. Moreover, there is a shortage of analysis and comparison of the diverse emerging backdoor countermeasures. Therefore, this work aims to provide the NLP community with a timely review of backdoor attacks and countermeasures. According to the attackers' capability and affected stage of the LMs, the attack surfaces are formalized into four categorizations: attacking the pre-trained model with fine-tuning (APMF) or parameter-efficient fine-tuning (APMP), attacking the final model with training (AFMT), and attacking Large Language Models (ALLM). Thus, attacks under each categorization are combed. The countermeasures are categorized into two general classes: sample inspection and model inspection. Thus, we review countermeasures and analyze their advantages and disadvantages. Also, we summarize the benchmark datasets and provide comparable evaluations for representative attacks and defenses. Drawing the insights from the review, we point out the crucial areas for future research on the backdoor, especially soliciting more efficient and practical countermeasures.

cs.CR↗