Searcharxiv⌕ Search

arXiv subjects

Xinkui Zhao

Publications and source records attributed to Xinkui Zhao.

At least 19 recordsLinked to original sources
arXiv2026

DuSE-Bench: Benchmarking Joint Reasoning over Evolving Dialogue and Documents

Persistent Web information assistants must combine evolving document facts with user requirements established and revised across sessions. Yet systematic evaluation of how changes in both sources affect response requirements and evidence attribution remains limited. To address this gap, we introduce Dual-Source Evolution Benchmark (DuSE-Bench), a benchmark comprising 1,000 tasks across five scenarios, including 700 whose required premises originate in both dialogue and documents. An evolving Support Graph guides task construction and preserves links from updated facts and reused conclusions to original turns and passages. Task annotations distinguish grounding source from state dependence and enable separate evaluation of response content, response type, and source attribution. Experiments across 11 baseline configurations reveal persistent challenges in joint reasoning over evolving dialogue and documents, with content accuracy on jointly grounded tasks reaching at most 56.7%. Moreover, improvements in content accuracy can coincide with poorer response choice, while content-correct responses frequently lack complete attribution to original evidence. Matched evidence substitutions further show that the effect of replacing one source's evidence depends on support from the other. These findings motivate coordinating dialogue memory and document retrieval to apply current user requirements and document facts together while preserving their original support. The data are available at https://github.com/JianfengGao7/DuSE-Bench.

cs.CL↗
arXiv2026

TenonAgent: Virtualizing Execution Residency for Memory-Oversubscribed LLM Agents

LLM agents execute as long-lived, multi-stage workflows with substantial resource-demand variation across phases. Provisioning each agent or sandbox for its lifetime peak leads to low resource utilization, particularly for memory, whereas aggressive oversubscription risks memory contention, out-of-memory failures, and costly workflow retries. This tension limits the concurrency and efficiency of agent serving under constrained host memory. We present TenonAgent, a runtime system that virtualizes sandbox residency by decoupling the logical agent's lifetime from its physical execution environment. To manage memory oversubscription, TenonAgent employs a dual-timescale control plane. At action boundaries, it combines predictive admission control with a fail-closed, restart-safe contract, dynamically materializing or dematerializing sandboxes based on global memory pressure. During active tool execution, a sched_ext-based kernel gate paces sandbox execution permissions to reduce unexpected concurrent memory growth. The gate provides feedback control over future execution opportunities. Evaluations across diverse agentic workloads show that, under the tested workloads and parameter settings, TenonAgent substantially reduces observed OOM failures and achieves a 100% task success rate in bursts reaching an 8.6x memory oversubscription ratio. Overall, it delivers up to a 5.5x speedup over unmanaged execution, outperforms existing agent-level and cgroup-based schedulers by 5x to 37x, and provides consistent benefits across both Docker and Firecracker microVM backends.

cs.OS↗
arXiv2026

SOV-CAD: Stepwise Orthographic Views Guided CAD Modeling Sequence Reconstruction

Reconstructing Computer-Aided Design (CAD) modeling sequences from images is crucial for preserving design intent and supporting parametric editing. However, existing methods typically generate full CAD sequences holistically, overlooking the iterative, feedback-driven nature of human design workflows. We address this limitation by introducing the rich stepwise visual supervision: at each modeling step, the system observes the target's orthographic projections, the projections of the incrementally constructed model, and the active sketch, enabling informed action selection. To effectively leverage this on-the-fly feedback, we propose SOV-CAD, a framework that formulates CAD reconstruction as a sequential decision-making task and employs offline reinforcement learning with a Decision Transformer architecture. This design incorporates continuous visual feedback guided by geometric alignment rewards, resulting in a more accurate and human-like modeling process. Extensive experiments show that SOV-CAD surpasses state-of-the-art methods in CAD sequence reconstruction while exhibiting strong data efficiency. Code of SOV-CAD is available at: https://github.com/LukePhong/SOV-CAD

cs.CV↗
arXiv2026

Dependency- and Layer-Aware Microservice Workflow Offloading and Service Image Caching for Edge Environments

The microservice architecture has been applied broadly in many mainstream computing environments. As one of the most prevalent environments, edge computing also widely employs microservices to handle diverse requests and tasks. In practical scenarios, microservices usually constitute workflows that are built based on service dependencies to execute tasks. This offers an opportunity to explore the offloading technology to better harness resources and accelerate task execution. Unfortunately, this problem has not been paid attention to by existing research, and thus some valuable resources (e.g., microservice image cache) remain obscure. To fill this gap, we innovatively study the problem of joint optimization for microservice workflow offloading and service image caching in edge. This problem is challenging due to several issues such as the complexity of its solution space, intricate relationships among shared image layers, long-range dependencies between workflow tasks, and the absence of a real-world collection of service images. To address these issues, this paper proposes an innovative dependency- and layer-aware workflow offloading and image caching framework for microservices. We collected a real-world collection of microservice image layer data and published both this collection and experimental codes on GitHub. Extensive results demonstrate that our framework achieves superior performances, for example, a 22. 38\% reduction in average task completion time compared to baselines and significantly-increased image hit rates.

cs.DC↗
arXiv2026

Shiva-DiT: Residual-Based Differentiable Top-$k$ Selection for Efficient Diffusion Transformers

Diffusion Transformers (DiTs) are costly at high resolution because self-attention scales quadratically with token sequence length. Existing pruning methods do not jointly provide end-to-end learnability, low training overhead, and deterministic token counts for predictable token-dependent computation. We propose Shiva-DiT, based on Residual-Based Differentiable Top-k Selection. Its forward pass executes hard top-k selection, while a residual-aware straight-through estimator propagates gradients to both token scores and the budget k without evaluating a second backbone path. A Context-Aware Router and Adaptive Ratio Policy learn layer- and timestep-dependent retention schedules under a target average budget. Experiments on SD3-Medium, Flux.1-dev, and PixArt-Σ show consistent reductions in FLOPs and measured latency. On SD3-Medium, Shiva-DiT provides four fidelity-latency operating points and reaches a 1.54x wall-clock speedup with competitive fidelity.

cs.LG↗
arXiv2026

Prediction-Robust Service Deployment with Capacity-Aware Edge Admission

Edge platforms instantiate executable services close to users to reduce request-serving cost, but each instance incurs a one-time deployment cost and remains useful only for a finite time-to-live (TTL). The resulting online decision is both prediction-sensitive and capacity-coupled: an optimistic forecast can waste deployment cost, whereas a delayed decision misses the burst it is intended to serve. We study this problem under a common TTL cost model and propose CAPSUM, a capacity-aware admission policy with an elastic specialization, CAPSUM-E. In the local elastic setting, every node-service trace is exactly a variable-price Bahncard instance. This reduction lets CAPSUM-E inherit PFSUM's tight prediction-error-dependent ratio, including $2/(1+β)$ consistency and $1/β$ robustness for $β>0$. A redirect-aware variant preserves the same local deployment schedule. For finite-capacity nodes, CAPSUM combines size-scaled break-even tests, a utilization-dependent shadow price, and evidence-density eviction; we prove capacity feasibility, scale invariance, and exact agreement with CAPSUM-E under an elastic configuration. We implement an exact local offline dynamic program and compare against direct common-model baselines and documented source-derived adapters for EDP-A, OREO, and uEDC-L. Experiments cover controlled prediction error, three synthetic demand regimes, a causal predictor on a public Globus Compute trace, and joint scaling to 1,024 nodes and 10,000 services. Under the common model, CAPSUM reduces normalized cost by 33.7-42.9% relative to the best source-derived adapter across the synthetic regimes and by 45.5% on the sampled trace.

cs.DC↗
arXiv2026

MUSE: A Heterogeneity-Aware Multimedia Search Engine for Mobile SoCs

On-device multimedia retrieval is vital for smartphones, enabling applications like cross-modal semantic search and multimodal personal AI agents. However, realizing efficient retrieval on mobile SoCs remains a critical systems challenge. Unlike text-based search, multimedia applications depend on high-dimensional cross-modal embeddings (often exceeding 1024 dimensions) and continuously expanding media streams (e.g., video lifelogs and ambient audio). Mobile devices must process these intensive workloads under strict latency, energy, and memory constraints. Existing vector retrieval systems, designed primarily for static server datasets, are poorly suited for the dynamic execution models of mobile SoCs, where interactive queries must coexist with continuous background media ingestion and index maintenance. To address these challenges, we propose MUSE, a MUltimedia Search Engine comprehensively co-designed for mobile SoCs. MUSE tackles these issues through an architecture-index co-design. It introduces a three-stage asynchronous execution pipeline that overlaps DMA transfers, NPU matrix computation, and hardware-vector data adaptation to overcome the memory wall of high-dimensional embeddings. Furthermore, it redesigns the Inverted File Index (IVF) and introduces workload-aware heterogeneous scheduling across the CPU, GPU, and NPU for hybrid interactive-and-ingestion workloads. Evaluated on Snapdragon 8-series SoCs using real-world multimodal datasets, MUSE improves query throughput by up to 1.4x at matched recall, achieves up to 7x faster index construction, and delivers up to 6x higher insertion throughput under concurrent streaming. Crucially, its accelerator-native design ensures strict physical compliance, capping peak device temperatures at 38°C and reducing total energy consumption by up to 4.6x compared to CPU-bound baselines.

cs.DC↗
arXiv2026

Better Call Grep: Evaluating and Improving Grep-Like Lexical Retrieval for Repository-Level Code Completion

Repository-level code completion remains challenging for large language models (LLMs) due to cross-file dependencies and limited context windows. Prior work addresses this challenge using Retrieval-Augmented Generation (RAG) frameworks based on semantic indexing or structure-aware graph analysis, but these approaches incur substantial computational overhead for index construction and maintenance. Motivated by common developer workflows that rely on lightweight search utilities (e.g., ripgrep), we revisit a fundamental yet underexplored question: how far can simple, index-free lexical retrieval support repository-level code completion before more complex retrieval mechanisms become necessary? To answer this question, we systematically investigate lightweight, index-free, intent-aware lexical retrieval through extensive empirical analysis. We first introduce Naive GrepRAG, a baseline framework in which LLMs autonomously generate ripgrep commands to retrieve relevant context. Despite its simplicity, Naive GrepRAG achieves performance comparable to sophisticated graph-based baselines. Further analysis shows that its effectiveness stems from retrieving lexically precise code fragments that are spatially closer to the completion site. We also identify key limitations of lexical retrieval, including sensitivity to noisy matches from high-frequency ambiguous keywords and context fragmentation caused by rigid truncation boundaries. To address these issues, we propose GrepRAG, which augments lexical retrieval with a lightweight post-processing pipeline featuring identifier-weighted re-ranking and structure-aware deduplication. Extensive evaluation on CrossCodeEval and RepoEval-Updated demonstrates that GrepRAG consistently outperforms state-of-the-art (SOTA) methods, achieving 7.04-15.58 percent relative improvement in code exact match (EM) over the best baseline on CrossCodeEval.

cs.SE↗
arXiv2026

FAVA: Formal Authorization for Verified Agents with Evidence-Backed Permission Graphs

Large language model (LLM) agents autonomously interleave semantic reasoning with complex system operations. In these dynamic environments, static tool-level permissions are fundamentally insufficient; safe authorization is highly context-dependent and heavily reliant on evolving runtime states and data flows. We present FAVA (Formal Authorization for Verified Agents), a permission-carrying authorization framework for agent execution. FAVA utilizes an LLM-guided Permission Intermediate Representation (IR) to translate ambiguous natural-language tasks into structured constraints. A deterministic lowering pass then converts this IR into an evidence-backed permission graph that explicitly tracks data flows, dependencies, and contextual labels. To provide strict security guarantees, a Satisfiability Modulo Theories (SMT) authorizer mathematically verifies the current graph against security policies before any effectful action executes. A runtime gateway then enforces the solver's result, either authorizing the execution or intercepting it with a precise counterexample. We evaluate FAVA across OpenAgentSafety, OctoBench, and ActPlane scenarios. Our evaluation demonstrates that FAVA achieves a 90.5% Decision Compliance Rate (DCR) over the aggregate dataset, successfully intercepting dynamic violating traces in the evaluated trace-conditioned scenarios.

cs.CR↗
arXiv2026

ViSAGE: Constructing Self-Correcting Memories for Long-Form Video Understanding

Multimodal agents operating in long-horizon environments must build and continually update multimedia memories to support entity-consistent, temporally grounded reasoning. However, existing agentic memory approaches often discard fine-grained dentity cues under aggressive compression and segment-wise processing. They also rely heavily on vector similarity retrieval, which can surface semantically related yet identity-mismatched evidence, leading to entity confusion, error propagation, and hallucinated answers. We propose ViSAGE, a multimodal agentic memory framework that constructs self-correcting, entity-centric memories. Specifically, ViSAGE anchors entity identity via cross-modal binding over long temporal ranges. It then applies bidirectional memory refinement to propagate delayed identity evidence, retroactively unifying historical records and improving future reasoning. We also introduce multi-agent cross-verification to assess retrieved evidence under an identity-evidence alignment onstraint, enabling abstention instead of unsupported answers when evidence is missing. Extensive results demonstrate that ViSAGE consistently outperforms the strongest baseline, achieving 5.9% higher accuracy.

cs.AI↗
arXiv2026

Industrial Data-Service-Knowledge Governance: Toward Integrated and Trusted Intelligence

The convergence of artificial intelligence, cyber-physical systems, and distributed networking has accelerated the evolution of industrial intelligence across edge, cloud, and cross-organizational communication environments. However, existing governance mechanisms remain fragmented across data management, service orchestration, and knowledge-based decision-making, making it difficult to ensure reliability, accountability, compliance, and explainability throughout the industrial intelligence stack. To address this gap, we present TRISK (TRusted Industrial Data-Service-Knowledge governance), a conceptual and taxonomic framework for trustworthy industrial intelligence. TRISK is grounded in a five-dimensional trust model covering quality, security, privacy, fairness, and explainability, and formalizes how trust is constructed, propagated, aggregated, and fed back across data, service, and knowledge layers in networked industrial systems. Through a structured synthesis of more than 100 representative studies, standards, and technical reports, we examine data governance as the foundation of trust construction, service governance as the mediation layer for trustworthy execution, and knowledge governance as the semantic anchor for reasoning, validation, and feedback adaptation. We further discuss industrial implementation patterns, cross-industry implications, and the role of emerging communication and computing technologies. Finally, we outline a future research roadmap toward adaptive, verifiable, and human-aligned industrial governance for Industry 5.0.

cs.CE↗
arXiv2026

Generation Quality-Latency Tradeoff-Aware Inference Offloading for Multimodal LLMs in Cloud-Edge Continuum

Beyond pure cloud, some efforts are being made to deploy Large Language Models (LLMs) in edge to accelerate inference response. So the deployment of LLMs in cloud-edge continuum becomes a promising paradigm, where the tasks involving multimodal data occupy a large part of requests. Under this continuum, users usually concern about multiple Quality-of-Service (QoS) attributes, but it is always intractable to jointly optimize them. In this paper, we propose to study the joint optimization of those attributes and focus on two key representatives, i.e., content generation quality and response latency. We propose to study the offloading technology to achieve a tradeoff between the two objectives in the cloud-edge collaborative Multimodal LLM (MLLM) system. However, it is highly difficult to predict generation quality and inference latency for MLLM inference tasks while optimizing this offloading process. To address these unprecedented difficulties, we propose a Quality-Latency Tradeoff-Aware MLLM Inference Offloading (QLMIO) framework to make decisions that optimally balance generation quality and response latency. Meanwhile, recognizing the absence of publicly available datasets tailored to the MLLM inference offloading problem, we constructed a real-world cloud-edge collaborative MLLM system and subsequently collected an MLLM Inference Offloading Benchmark (MIOBench) to comprehensively evaluate our framework and facilitate the study of this problem. Extensive experimental results demonstrate that the QLMIO framework reduces latency by up to 58.14\% compared to baselines, while simultaneously matching the task completion rate achieved under the case that executes all requests exclusively on a cloud server. The dataset and codes are available at Github.

cs.DC↗
arXiv2026

Measurement-Driven Diagnosis and Mitigation of Host-CPU Co-location Interference in Single-GPU LLM Serving on a Multi-GPU Server

Host CPUs in GPU servers are often under-used during LLM inference. Co-locating CPU workloads can improve resource use, but it can also seriously hurt serving quality. Existing work mainly improves LLM serving engines or studies CPU-GPU boundary delays. It gives limited guidance on how external CPU workloads affect the serving path and how operators should choose protection policies. This paper studies host-CPU co-location interference in single-GPU LLM serving. We show that the main observed problem is not slower GPU kernels. Instead, CPU workloads amplify long tails in CPU-side serving stages before GPU work is submitted. To capture this effect, we introduce the Core Path Tail Index (CPTI) and Core Tail Suppression (CTS). Based on these metrics, we build CoTail, a measurement-driven diagnostic procedure that screens workload risk, profiles serving-stage tails, selects OS-level protections, and validates decode SLO compliance. In our primary setup, unprotected nginx co-location reduces throughput by 78.8%, increases TTFT by 429.5%, and increases TPOT by 362.4%. CoTail-guided protections improve nginx throughput by up to 4.4x and reduce TPOT by 4.5x. Under a common-baseline deployment SLO, CoTail satisfies all 12 oracle-feasible held-out cases, compared with 10/12 for Always-rt and 11/12 for Macro-only. It also reduces RT usage from 28 to 22 cases and lowers mean co-tenant slowdown from 56.65% to 51.21%.

cs.DC↗
arXiv2026

ATOM: Instantiating Budget-Controllable Multi-Agent Collaboration via Nucleus-Electron Hierarchy

Large Language Model (LLM)-based multi-agent systems rely on optimized collaboration topologies to balance performance and communication costs. However, current methods struggle with the inherent stability-extensibility trade-off and often misalign computational budgets with query difficulty. We propose \textsc{ATOM}, an adaptive framework that generates budget-controllable collaboration graphs via a novel task-driven reinforcement learning paradigm. Inspired by atomic structures, \textsc{ATOM} employs a nucleus-electron hierarchy: it maintains a stable, offline-learned collaboration backbone (the nucleus) while dynamically activating query-conditioned agents (electrons) during inference. Crucially, a complexity-aware budgeting strategy aligns resource consumption with task demands by estimating query difficulty to strictly regulate electron instantiation. Extensive experiments across six diverse benchmarks demonstrate that \textsc{ATOM} achieves state-of-the-art performance while improving token efficiency by up to $30\%$ compared to strong baselines.

cs.MA↗
arXiv2026

DRAMA: Next-Gen Dynamic Orchestration for Resilient Multi-Agent Ecosystems in Flux

Multi-agent systems (MAS) have demonstrated significant effectiveness in addressing complex problems through coordinated collaboration among heterogeneous agents. However, real-world environments and task specifications are inherently dynamic, characterized by frequent changes, uncertainty, and variability. Despite this, most existing MAS frameworks rely on static architectures with fixed agent capabilities and rigid task allocation strategies, which greatly limits their adaptability to evolving conditions. This inflexibility poses substantial challenges for sustaining robust and efficient multi-agent cooperation in dynamic and unpredictable scenarios. To address these limitations, we propose DRAMA: a Dynamic and Robust Allocation-based Multi-Agent System designed to facilitate resilient collaboration in rapidly changing environments. DRAMA features a modular architecture with a clear separation between the control plane and the worker plane. Both agents and tasks are abstracted as resource objects with well-defined lifecycles, while task allocation is achieved via an affinity-based, loosely coupled mechanism. The control plane enables real-time monitoring and centralized planning, allowing flexible and efficient task reassignment as agents join, depart, or become unavailable, thereby ensuring continuous and robust task execution. The worker plane comprises a cluster of autonomous agents, each with local reasoning, task execution, the ability to collaborate, and the capability to take over unfinished tasks from other agents when needed.

cs.MA↗
arXiv2026

CoreGuard: Safeguarding Foundational Capabilities of LLMs Against Model Stealing in Edge Deployment

Proprietary large language models (LLMs) exhibit strong generalization capabilities across diverse tasks and are increasingly deployed on edge devices for efficiency and privacy reasons. However, deploying proprietary LLMs at the edge without adequate protection introduces critical security threats. Attackers can extract model weights and architectures, enabling unauthorized copying and misuse. Even when protective measures prevent full extraction of model weights, attackers may still perform advanced attacks, such as fine-tuning, to further exploit the model. Existing defenses against these threats typically incur significant computational and communication overhead, making them impractical for edge deployment. To safeguard the edge-deployed LLMs, we introduce CoreGuard, a computation- and communication-efficient protection method. CoreGuard employs an efficient protection protocol to reduce computational overhead and minimize communication overhead via a propagation protocol. Extensive experiments show that CoreGuard achieves upper-bound security protection with negligible overhead.

cs.CR↗
arXiv2026

Reliable Microservice Tail Latency Prediction via Decoupled Dual-Stream Learning and Gradient Modulation

Microservice architectures enable scalable cloud-native applications; however, the distributed nature of these systems complicates the maintenance of strict Service Level Objectives. Accurately predicting window-level P95 tail latency remains difficult due to the complex interactions between software workload propagation and infrastructure resource limits. Existing predictive models struggle to capture these dynamics because the lack of explicit separation between traffic metrics and resource metrics causes misaligned feature representations. Building on this suboptimal data treatment, the unified architectures of prior approaches fail to isolate cascading service dependencies from localized processing capacity. Due to this entanglement, joint training suffers from an optimization imbalance wherein resource features converge faster and dominate gradient updates, thereby preventing the learning of underlying software topologies. To address these challenges, we propose USRFNet, a dual-stream framework that separates the modeling of demand and capacity. The proposed framework utilizes a Graph Neural Network to model the spatial interactions of traffic workloads across software-level service dependencies, and a gating MLP to independently extract infrastructure-level resource dynamics. The model then integrates these representations through hierarchical tensor fusion. To resolve the training imbalance, we introduce a Reliability-Aware Gradient Modulation strategy that dynamically rescales gradients based on the generalization ratio of each data stream. Experiments on three large-scale real-world benchmarks demonstrate that USRFNet outperforms state-of-the-art methods in prediction accuracy. Specifically, compared to the best-performing baselines, the proposed framework achieves relative MAPE reductions ranging from 15.62% to 26.11% across the evaluated datasets.

cs.LG↗
arXiv2026

SortingHat: Redefining Operating Systems Education with a Tailored Digital Teaching Assistant

Operating Systems (OS) courses are among the most challenging in computer science education due to the complexity of internal structures and the diversity of running environments. Traditional teaching methods often fail to address the diverse backgrounds, learning speeds, and practical needs of students. To tackle these challenges, we present SortingHat, a personalized digital teaching assistant tailored specifically for OS education. SortingHat integrates advanced AI technologies, including a retrieval augmented generation (RAG) framework and multi agent reinforcement learning (MARL), to deliver adaptive, scalable, and effective educational support. SortingHat features a 3D digital human interface powered by large language models (LLMs) to provide personalized, empathetic, and context aware guidance. It generates tailored exercises based on each student's learning history and academic performance, reinforcing weak areas and challenging advanced concepts. Additionally, the system incorporates a robust evaluation pipeline that ensures fair, consistent, and unbiased grading of student submissions while delivering personalized, actionable feedback for improvement. By combining personalized guidance, adaptive content creation, and automated assessment, SortingHat transforms OS education into an engaging, immersive, and scalable experience.

cs.HC↗