arXiv subjects

Dingyu Yang

Publications and source records attributed to Dingyu Yang.

At least 19 recordsLinked to original sources

arXiv2026

Token Economics for LLM Agents: A Dual-View Study from Computing and Economics

As LLM agents evolve, tokens have emerged as the core economic primitives of Agentic AI. However, their exponential consumption introduces severe computational, collaborative, and security bottlenecks. Current surveys remain fragmented across system optimization, architecture design, and trust, lacking a unified framework to evaluate the fundamental trade-off between output quality and economic cost. To bridge this gap, this survey presents the first comprehensive survey of Token Economics. By unifying computer science and economics, we conceptualize tokens as production factors, exchange mediums, and units of account. We synthesize existing literature across a four-dimensional taxonomy: (1) Micro-level (Single Agent): Optimizing budget-constrained factor substitution via neoclassical firm theory. (2) Meso-level (Multi-Agent Systems): Minimizing collaboration friction using transaction cost and principal-agent theories. (3) Macro-level (Agent Ecosystems): Addressing congestion externalities and pricing via mechanism design. (4) Security: Internalizing adversarial threats as endogenous economic constraints. Finally, we outline frontier directions, including differentiable token budgets and dynamic markets, to lay the theoretical foundation for scalable next-generation agent systems.

cs.AI↗

arXiv2026

Hestia: Hyperthread-Level Scheduling for Cloud Microservices with Interference-Aware Attention

Modern cloud servers routinely co-locate multiple latency-sensitive microservice instances to improve resource efficiency. However, the diversity of microservice behaviors, coupled with mutual performance interference under simultaneous multithreading (SMT), makes large-scale placement increasingly complex. Existing interference aware schedulers and isolation techniques rely on coarse core-level profiling or static resource partitioning, leaving asymmetric hyperthread-level heterogeneity and SMT contention dynamics largely unmodeled. We present Hestia, a hyperthread-level, interference-aware scheduling framework powered by self-attention. Through an extensive analysis of production traces encompassing 32,408 instances across 3,132 servers, we identify two dominant contention patterns -- sharing-core (SC) and sharing-socket (SS) -- and reveal strong asymmetry in their impact. Guided by these insights, Hestia incorporates (1) a self-attention-based CPU usage predictor that models SC/SS contention and hardware heterogeneity, and (2) an interference scoring model that estimates pairwise contention risks to guide scheduling decisions. We evaluate Hestia through large-scale simulation and a real production deployment. Hestia reduces the 95th-percentile service latency by up to 80\%, lowers overall CPU consumption by 2.3\% under the same workload, and surpasses five state-of-the-art schedulers by up to 30.65\% across diverse contention scenarios.

cs.DC↗

arXiv2026

SVFusion: A CPU-GPU Co-Processing Architecture for Large-Scale Real-Time Vector Search

Approximate Nearest Neighbor Search (ANNS) underpins modern applications such as information retrieval and recommendation. With the rapid growth of vector data, efficient indexing for real-time vector search has become rudimentary. Existing CPU-based solutions support updates but suffer from low throughput, while GPU-accelerated systems deliver high performance but face challenges with dynamic updates and limited GPU memory, resulting in a critical performance gap for continuous, large-scale vector search requiring both accuracy and speed. In this paper, we present SVFusion, a GPU-CPU-disk collaborative framework for real-time vector search that bridges sophisticated GPU computation with online updates. SVFusion leverages a hierarchical vector index architecture that employs CPU-GPU co-processing, along with a workload-aware vector caching mechanism to maximize the efficiency of limited GPU memory. It further enhances performance through real-time coordination with CUDA multi-stream optimization and adaptive resource management, along with concurrency control that ensures data consistency under interleaved queries and updates. Empirical results demonstrate that SVFusion achieves significant improvements in query latency and throughput, exhibiting a 20.9x higher throughput on average and 1.3x to 50.7x lower latency compared to baseline methods, while maintaining high recall for large-scale datasets under various streaming workloads.

cs.DB↗

arXiv2026

SafeLoad: Efficient Admission Control Framework for Identifying Memory-Overloading Queries in Cloud Data Warehouses

Memory overload is a common form of resource exhaustion in cloud data warehouses. When database queries fail due to memory overload, it not only wastes critical resources such as CPU time but also disrupts the execution of core business processes, as memory-overloading (MO) queries are typically part of complex workflows. If such queries are identified in advance and scheduled to memory-rich serverless clusters, it can prevent resource wastage and query execution failure. Therefore, cloud data warehouses desire an admission control framework with high prediction precision, interpretability, efficiency, and adaptability to effectively identify MO queries. However, existing admission control frameworks primarily focus on scenarios like SLA satisfaction and resource isolation, with limited precision in identifying MO queries. Moreover, there is a lack of publicly available MO-labeled datasets with workloads for training and benchmarking. To tackle these challenges, we propose SafeLoad, the first query admission control framework specifically designed to identify MO queries. Alongside, we release SafeBench, an open-source, industrial-scale benchmark for this task, which includes 150 million real queries. SafeLoad first filters out memory-safe queries using the interpretable discriminative rule. It then applies a hybrid architecture that integrates both a global model and cluster-level models, supplemented by a misprediction correction module to identify MO queries. Additionally, a self-tuning quota management mechanism dynamically adjusts prediction quotas per cluster to improve precision. Experimental results show that SafeLoad achieves state-of-the-art prediction performance with low online and offline time overhead. Specifically, SafeLoad improves precision by up to 66% over the best baseline and reduces wasted CPU time by up to 8.09x compared to scenarios without SafeLoad.

cs.DB↗

arXiv2026

DeXOR: Enabling XOR in Decimal Space for Streaming Lossless Compression of Floating-point Data

With streaming floating-point numbers being increasingly prevalent, effective and efficient compression of such data is critical. Compression schemes must be able to exploit the similarity, or smoothness, of consecutive numbers and must be able to contend with extreme conditions, such as high-precision values or the absence of smoothness. We present DeXOR, a novel framework that enables decimal XOR procedure to encode decimal-space longest common prefixes and suffixes, achieving optimal prefix reuse and effective redundancy elimination. To ensure accurate and low-cost decompression even with binary-decimal conversion errors, DeXOR incorporates 1) scaled truncation with error-tolerant rounding and 2) different bit management strategies optimized for decimal XOR. Additionally, a robust exception handler enhances stability by managing floating-point exponents, maintaining high compression ratios under extreme conditions. In evaluations across 22 datasets, DeXOR surpasses state-of-the-art schemes, achieving a 15% higher compression ratio and a 20% faster decompression speed while maintaining a competitive compression speed. DeXOR also offers scalability under varying conditions and exhibits robustness in extreme scenarios where other schemes fail.

cs.DB↗

arXiv2025

EDGC: Entropy-driven Dynamic Gradient Compression for Efficient LLM Training

Training large language models (LLMs) poses significant challenges regarding computational resources and memory capacity. Although distributed training techniques help mitigate these issues, they still suffer from considerable communication overhead. Existing approaches primarily rely on static gradient compression to enhance communication efficiency; however, these methods neglect the dynamic nature of evolving gradients during training, leading to performance degradation. Accelerating LLM training via compression without sacrificing performance remains a challenge. In this paper, we propose an entropy-driven dynamic gradient compression framework called EDGC. The core concept is to adjust the compression rate during LLM training based on the evolving trends of gradient entropy, taking into account both compression efficiency and error. EDGC consists of three key components.First, it employs a down-sampling method to efficiently estimate gradient entropy, reducing computation overhead. Second, it establishes a theoretical model linking compression rate with gradient entropy, enabling more informed compression decisions. Lastly, a window-based adjustment mechanism dynamically adapts the compression rate across pipeline stages, improving communication efficiency and maintaining model performance. We implemented EDGC on a 32-NVIDIA-V100 cluster and a 64-NVIDIA-H100 cluster to train GPT2-2.5B and GPT2-12.1B, respectively. The results show that EDGC significantly reduces communication latency and training time by up to 46.45% and 16.13% while preserving LLM accuracy.

cs.LG↗

arXiv2025

GFS: A Preemption-aware Scheduling Framework for GPU Clusters with Predictive Spot Instance Management

The surge in large language models (LLMs) has fundamentally reshaped the landscape of GPU usage patterns, creating an urgent need for more efficient management strategies. While cloud providers employ spot instances to reduce costs for low-priority (LP) tasks, existing schedulers still grapple with high eviction rates and lengthy queuing times. To address these limitations, we present GFS, a novel preemptive scheduling framework that enhances service-level objective (SLO) compliance for high-priority (HP) tasks while minimizing preemptions to LP tasks. Firstly, GFS utilizes a lightweight forecasting model that predicts GPU demand among different tenants, enabling proactive resource management. Secondly, GFS employs a dynamic allocation mechanism to adjust the spot quota for LP tasks with guaranteed durations. Lastly, GFS incorporates a preemptive scheduling policy that prioritizes HP tasks while minimizing the impact on LP tasks. We demonstrate the effectiveness of GFS through both real-world implementation and simulations. The results show that GFS reduces eviction rates by 33.0\%, and cuts queuing delays by 44.1\% for LP tasks. Furthermore, GFS enhances the GPU allocation rate by up to 22.8\% in real production clusters. In a production cluster of more than 10,000 GPUs, GFS yields roughly \$459,715 in monthly benefits.

cs.DC↗

arXiv2025

Atys: An Efficient Profiling Framework for Identifying Hotspot Functions in Large-scale Cloud Microservices

To handle the high volume of requests, large-scale services are comprised of thousands of instances deployed in clouds. These services utilize diverse programming languages and are distributed across various nodes as encapsulated containers. Given their vast scale, even minor performance enhancements can lead to significant cost reductions. In this paper, we introduce Atys1, an efficient profiling framework specifically designed to identify hotspot functions within large-scale distributed services. Atys presents four key features. First, it implements a language-agnostic adaptation mechanism for multilingual microservices. Second, a two-level aggregation method is introduced to provide a comprehensive overview of flamegraphs. Third, we propose a function selective pruning (FSP) strategy to enhance the efficiency of aggregating profiling results. Finally, we develop a frequency dynamic adjustment (FDA) scheme that dynamically modifies sampling frequency based on service status, effectively minimizing profiling cost while maintaining accuracy. Cluster-scale experiments on two benchmarks show that the FSP strategy achieves a 6.8% reduction in time with a mere 0.58% mean average percentage error (MAPE) in stack traces aggregation. Additionally, the FDA scheme ensures that the mean squared error (MSE) remains on par with that at high sampling rates, while achieving an 87.6% reduction in cost.

cs.PF↗

arXiv2025

LKD-KGC: Domain-Specific KG Construction via LLM-driven Knowledge Dependency Parsing

Knowledge Graphs (KGs) structure real-world entities and their relationships into triples, enhancing machine reasoning for various tasks. While domain-specific KGs offer substantial benefits, their manual construction is often inefficient and requires specialized knowledge. Recent approaches for knowledge graph construction (KGC) based on large language models (LLMs), such as schema-guided KGC and reference knowledge integration, have proven efficient. However, these methods are constrained by their reliance on manually defined schema, single-document processing, and public-domain references, making them less effective for domain-specific corpora that exhibit complex knowledge dependencies and specificity, as well as limited reference knowledge. To address these challenges, we propose LKD-KGC, a novel framework for unsupervised domain-specific KG construction. LKD-KGC autonomously analyzes document repositories to infer knowledge dependencies, determines optimal processing sequences via LLM driven prioritization, and autoregressively generates entity schema by integrating hierarchical inter-document contexts. This schema guides the unsupervised extraction of entities and relationships, eliminating reliance on predefined structures or external knowledge. Extensive experiments show that compared with state-of-the-art baselines, LKD-KGC generally achieves improvements of 10% to 20% in both precision and recall rate, demonstrating its potential in constructing high-quality domain-specific KGs.

cs.CL↗

arXiv2025

PromCopilot: Simplifying Prometheus Metric Querying in Cloud Native Online Service Systems via Large Language Models

With the increasing complexity of modern online service systems, understanding the state and behavior of the systems is essential for ensuring their reliability and stability. Therefore, metric monitoring systems are widely used and become an important infrastructure in online service systems. Engineers usually interact with metrics data by manually writing domain-specific language (DSL) queries to achieve various analysis objectives. However, writing these queries can be challenging and time-consuming, as it requires engineers to have high programming skills and understand the context of the system. In this paper, we focus on PromQL, which is the metric query DSL provided by the widely used metric monitoring system Prometheus. We aim to simplify metrics querying by enabling engineers to interact with metrics data in Prometheus through natural language, and we call this task text-to-PromQL. Building upon the insight, this paper proposes PromCopilot, a Large Language Model-based text-to-PromQL framework. PromCopilot first uses a knowledge graph to describe the complex context of a cloud native online service system. Then, through the synergistic reasoning of LLMs and the knowledge graph, PromCopilot transforms engineers' natural language questions into PromQL queries. To evaluate PromCopilot, we manually construct the first text-to-PromQL benchmark dataset which contains 280 metric query questions. The experiment results show that PromCopilot is effective in text-to-PromQL. When using GPT-4 as the backbone LLM, PromCopilot achieves an accuracy of 69.1\% in translating natural language questions to PromQL queries when using. To the best of our knowledge, this paper is the first study of text-to-PromQL, and PromCopilot pioneered the DSL generation framework for metric querying and analysis.

cs.SE↗

arXiv2025

Harnessing Feature Resonance under Arbitrary Target Alignment for Out-of-Distribution Node Detection

Detecting out-of-distribution (OOD) nodes in the graph-based machine-learning field is challenging, particularly when in-distribution (ID) node multi-category labels are unavailable. Thus, we focus on feature space rather than label space and find that, ideally, during the optimization of known ID samples, unknown ID samples undergo more significant representation changes than OOD samples, even if the model is trained to fit random targets, which we called the Feature Resonance phenomenon. The rationale behind it is that even without gold labels, the local manifold may still exhibit smooth resonance. Based on this, we further develop a novel graph OOD framework, dubbed Resonance-based Separation and Learning (RSL), which comprises two core modules: (i) a more practical micro-level proxy of feature resonance that measures the movement of feature vectors in one training step. (ii) integrate with synthetic OOD nodes strategy to train an effective OOD classifier. Theoretically, we derive an error bound showing the superior separability of OOD nodes during the resonance period. Extensive experiments on a total of thirteen real-world graph datasets empirically demonstrate that RSL achieves state-of-the-art performance.

cs.LG↗

arXiv2024

Mitigating Interference of Microservices with a Scoring Mechanism in Large-scale Clusters

Co-locating latency-critical services (LCSs) and best-effort jobs (BEJs) constitute the principal approach for enhancing resource utilization in production. Nevertheless, the co-location practice hurts the performance of LCSs due to resource competition, even when employing isolation technology. Through an extensive analysis of voluminous real trace data derived from two production clusters, we observe that BEJs typically exhibit periodic execution patterns and serve as the primary sources of interference to LCSs. Furthermore, despite occupying the same level of resource consumption, the diverse compositions of BEJs can result in varying degrees of interference on LCSs. Subsequently, we propose PISM, a proactive Performance Interference Scoring and Mitigating framework for LCSs through the optimization of BEJ scheduling. Firstly, PISM adopts a data-driven approach to establish a characterization and classification methodology for BEJs. Secondly, PISM models the relationship between the composition of BEJs on servers and the response time (RT) of LCSs. Thirdly, PISM establishes an interference scoring mechanism in terms of RT, which serves as the foundation for BEJ scheduling. We assess the effectiveness of PISM on a small-scale cluster and through extensive data-driven simulations. The experiment results demonstrate that PISM can reduce cluster interference by up to 41.5%, and improve the throughput of long-tail LCSs by 76.4%.

cs.DC↗

arXiv2024

Humas: A Heterogeneity- and Upgrade-aware Microservice Auto-scaling Framework in Large-scale Data Centers

An effective auto-scaling framework is essential for microservices to ensure performance stability and resource efficiency under dynamic workloads. As revealed by many prior studies, the key to efficient auto-scaling lies in accurately learning performance patterns, i.e., the relationship between performance metrics and workloads in data-driven schemes. However, we notice that there are two significant challenges in characterizing performance patterns for large-scale microservices. Firstly, diverse microservices demonstrate varying sensitivities to heterogeneous machines, causing difficulty in quantifying the performance difference in a fixed manner. Secondly, frequent version upgrades of microservices result in uncertain changes in performance patterns, known as pattern drifts, leading to imprecise resource capacity estimation issues. To address these challenges, we propose Humas, a heterogeneity- and upgrade-aware auto-scaling framework for large-scale microservices. Firstly, Humas quantifies the difference in resource efficiency among heterogeneous machines for various microservices online and normalizes their resources in standard units. Additionally, Humas develops a least squares density-difference (LSDD) based algorithm to identify pattern drifts caused by upgrades. Lastly, Humas generates capacity adjustment plans for microservices based on the latest performance patterns and predicted workloads. The experiment results conducted on 50 real microservices with over 11,000 containers demonstrate that Humas improves resource efficiency and performance stability by approximately 30.4% and 48.0%, respectively, compared to state-of-the-art approaches.

cs.DC↗

arXiv2024

MOPAR: A Model Partitioning Framework for Deep Learning Inference Services on Serverless Platforms

With its elastic power and a pay-as-you-go cost model, the deployment of deep learning inference services (DLISs) on serverless platforms is emerging as a prevalent trend. However, the varying resource requirements of different layers in DL models hinder resource utilization and increase costs, when DLISs are deployed as a single function on serverless platforms. To tackle this problem, we propose a model partitioning framework called MOPAR. This work is based on the two resource usage patterns of DLISs: global differences and local similarity, due to the presence of resource dominant (RD) operators and layer stacking. Considering these patterns, MOPAR adopts a hybrid approach that initially divides the DL model vertically into multiple slices composed of similar layers to improve resource efficiency. Slices containing RD operators are further partitioned into multiple sub-slices, enabling parallel optimization to reduce inference latency. Moreover, MOPAR comprehensively employs data compression and share-memory techniques to offset the additional time introduced by communication between slices. We implement a prototype of MOPAR and evaluate its efficacy using four categories of 12 DL models on OpenFaaS and AWS Lambda. The experiment results show that MOPAR can improve the resource efficiency of DLISs by 27.62\% on average, while reducing latency by about 5.52\%. Furthermore, based on Lambda's pricing, the cost of running DLISs is reduced by about 2.58 $\times$ using MOPAR.

cs.DC↗

arXiv2022

Ultra-narrowband terahertz circular dichroism driven by planar metasurface supporting chiral quasi bound states in continuum

Terahertz (THz) chirality pursues customizable manipulation from narrowband to broadband. While conventional THz chirality is restricted by non-negligible linewidth and unable to handle narrowband well. Recently, the concept "quasi bound states in continuum" (quasi-BIC) is introduced to optics resonance system whose the quality factor can be extremely high with the ultra-low radiative loss, thus providing a conceptual feasibility for wave control with ultra-narrow linewidth. Herein, we construct quasi-BIC in a planar all-silicon THz metasurface with in-plane C2 and mirror symmetries breaking. Such system not only exposes the symmetry-protected BIC, but also exposes the parameter-tuned BIC assigned to single resonance type. An extremely narrow linewidth (below 0.06 GHz) with high quality factor (104 level) is obtained at quasi-BIC frequency, which achieves the ultra-narrowband THz chirality.

physics.optics↗

arXiv2020

Distributed Processing of k Shortest Path Queries over Dynamic Road Networks

The problem of identifying the k-shortest paths (KSPs for short) in a dynamic road network is essential to many location-based services. Road networks are dynamic in the sense that the weights of the edges in the corresponding graph constantly change over time, representing evolving traffic conditions. Very often such services have to process numerous KSP queries over large road networks at the same time, thus there is a pressing need to identify distributed solutions for this problem. However, most existing approaches are designed to identify KSPs on a static graph in a sequential manner (i.e., the (i+1)-th shortest path is generated based on the i-th shortest path), restricting their scalability and applicability in a distributed setting. We therefore propose KSP-DG, a distributed algorithm for identifying k-shortest paths in a dynamic graph. It is based on partitioning the entire graph into smaller subgraphs, and reduces the problem of determining KSPs into the computation of partial KSPs in relevant subgraphs, which can execute in parallel on a cluster of servers. A distributed two-level index called DTLP is developed to facilitate the efficient identification of relevant subgraphs. A salient feature of DTLP is that it indexes a set of virtual paths that are insensitive to varying traffic conditions, leading to very low maintenance cost in dynamic road networks. This is the first treatment of the problem of processing KSP queries over dynamic road networks. Extensive experiments conducted on real road networks confirm the superiority of our proposal over baseline methods.

cs.DB↗

arXiv2018

Fukaya category of Landau-Ginzburg model

This article introduces and provides the mathematical foundation of the open string Floer theory of Landau-Ginzburg model viaWitten equation. We introduce the concept of regular tame exact Landau-Ginzburg system on a noncompact Kaehler manifold, and define the notion of Landau-Ginzburg branes, as the objects of the Fukaya category. The study on Witten equation in our context provides the construction of the Fukaya category of Landau-Ginzburg model which was conjectured to be existed in Gaiotto-Moore-Witten's work [22] and Kapranov-Kontsevich-Soibelman's work [30]. This is the first paper in the series of studies on Landau-Ginzburg models in the contexts of the mirror symmetry and other topics.

math.SG↗

arXiv2015

Virtual harmony

This article serves a few purposes. First of all, it reviews polyfold--Kuranishi correspondence I (http://arxiv.org/abs/1402.7008) and previews and samples some results from four papers I have been preparing. It is also a written-up and expanded version of a talk I gave at a symplectic conference in Chengdu on June 28, 2015, and it intends to provide bridges and compatibility between various pairs of virtual techniques and to demonstrate some unity among various technical viewpoints in the constructions of structures on moduli spaces in symplectic geometry. More precisely, the abstract perturbative structures (or interchangeably, virtual structures) present in each virtual theory discussed in this paper (and sometimes even the way they essentially originate in applications) are identified pairwise in a way that intertwines the (non-)perturbation mechanisms. To be more helpful to readers and not get them buried under technicalities and notations, we give the ideas and appropriate level of details so that the results will be clear to the relevant experts; meanwhile the ideas of each virtual machinery and how they are related should come through to more application-minded readers so that they might get encouraged to read papers on a given virtual machinery and possibly apply it to remove some technical assumptions in their results. It is meant to be a service to the symplectic community.

math.SG↗