arXiv subjects

Quan Zhou

Publications and source records attributed to Quan Zhou.

At least 19 recordsLinked to original sources

arXiv2026

DualGuard: Comprehensive and Efficient Memory Safety across Sequential and Speculative Execution

Although decades of research have yielded sophisticated mechanisms for enforcing memory safety, speculative execution can subvert these protections through transient memory operations. Conversely, defenses designed specifically for speculative safety, such as Speculative Load Hardening (SLH), cannot enforce memory safety during non-speculative execution. A naive combination of defenses from both domains raises two fundamental challenges: (1) which combination of otherwise fragmented defenses can ensure dual memory safety under speculative and non-speculative executions? (2) whether such combinations introduce redundant protections that incur unnecessary performance overhead? We present DualGuard, a unified framework for dual memory protection under non-speculative and speculative executions. DualGuard systematically analyzes memory behaviors, classifying them into four categories, and further identifies the minimal composition requirements for dual memory safety, which allow us to strategically combine existing mitigations and design new ones to avoid redundant or excessive protections. We implement DualGuard through a combination of static analysis and runtime enforcement within the LLVM infrastructure. Our evaluation demonstrates that DualGuard reduces overhead by 3%-72% compared to naive compositions of fragmented defenses. In some cases, it even incurs lower overhead than approaches that guarantee memory safety solely under either non-speculative or speculative execution.

cs.CR↗

arXiv2026

Learning Representative Port Subsets for Efficient Fluid Antenna Systems

Fluid antenna systems (FASs) provide many candidate ports, but probing all ports for each channel realization incurs substantial online overhead. We study a complementary offline problem: compress the candidate-port space once and reuse the compressed set across future channel realizations. The method collects optimal ports from sampled training realizations, requiring only one scan of the candidate ports per realization. On a DeepMIMO 7-by-7 FAS with 49 candidate ports, 120 training realizations yield 27 representative ports on average while retaining 99.995% of the channel power gain achieved by exhaustive port selection.

cs.IT↗

arXiv2026

Uniform ergodicity and spectral gap of importance-tempered MCMC with heavy-tailed targets

We study the importance-tempered random-walk Metropolis (RWM) algorithm on $\mathbb{R}^d$, which targets the tempered density $π_β(x) \propto π(x)^β$ with $β\in (0,1)$ and corrects for the resulting bias by importance weighting. The dynamics of this sampler can be described by a continuous-time Markov chain whose embedded chain is the tempered RWM. We prove that when the tails of $π$ decay at rate $\|x\|^{-γ}$, this continuous-time chain is uniformly ergodic if and only if $d/γ< β< (γ- 2)/γ$. Further, the global spectral gap of this chain can be controlled via the spectral gap of the chain restricted to a ball with radius growing polynomially in $d$. Our theory and simulation studies show that, with a suitable choice of $β$, importance tempering can substantially reduce sensitivity to initialization and significantly improve efficiency over the standard RWM.

stat.CO↗

arXiv2026

Re-emergence of fingering in solidifying liquid-metal drops

Solidification is expected to suppress fingering by removing mobile liquid. Here we report an unexpected re-emergence of fingering in impacting molten-tin drops: cooling first weakens peripheral fingers, but stronger undercooling restores them. At weak undercooling, fingers originate from a decelerating liquid rim and solidification reduces the effective acceleration that drives Rayleigh--Taylor-like amplification. At strong undercooling, solidification structures develop while liquid is still spreading, creating lateral mobility contrasts that redirect the residual flow. Comparison between the kinematically estimated solidification time and the experimentally measured fingering-onset and growth-termination times reveals a reversal in temporal ordering across the temperature range. A physics-based reduced-order model combines solidification-modified rim deceleration with mobility-gradient growth, captures the characteristic onset dynamics and accounts for the final finger length within the same physical framework.These findings establish solidification as an active means of regulating interfacial flow and morphology, with broad implications spanning metal-based additive manufacturing and debris mitigation in semiconducting industry.

physics.flu-dyn↗

arXiv2026

Multinomial Subset Routing with Sum-Max Rewards and Operational Constraints

We study online routing to subsets of experts under aggregate bandit feedback and long-run operational constraints. Expert contributions can be complementary: for each task dimension, the best selected expert determines the contribution, and the total reward aggregates these dimension-wise maxima. At the same time, capacity, budget, and fairness requirements impose lower and upper bounds on long-run expert activation frequencies. A fixed deterministic subset cannot generally satisfy such heterogeneous requirements, motivating a stochastic routing policy. We formalize this problem as Multinomial Subset Routing (MSR). The learner maintains a distribution $q$ over $K$ experts, samples an expert independently $M$ times from $q$, and routes each task to the distinct sampled experts. We propose OMD-Approachability, which combines online mirror descent with Blackwell's approachability to optimize MSR under two-sided operational constraints. We establish $O(T^{-1/2})$ average reward regret and expected constraint violation. We also quantify the approximation gap induced by a linear surrogate, and evaluate the approach on a real-world crowdsourcing dataset.

cs.LG↗

arXiv2026

A Closed-Loop Control Architecture for Reliable Constraint Satisfaction in LLM Text Generation

Software systems increasingly embed a large language model in features that must satisfy a numeric output constraint, that is, a requirement expressible as a number or an interval and checkable by code, such as a target word count or a target readability grade band. Because such a model is non-deterministic, is configured through natural-language instructions rather than a typed interface, and satisfies a stated requirement only approximately, a single prompt neither reliably meets the target nor preserves the source content. This paper presents and evaluates a closed-loop control architecture for this problem. It has five stages: generate, evaluate, adjust, archive, and analyze. The model is called only to write and to edit text, while deterministic code compares a composite readability value against a target band, rejects any edit that drops source entities, numbers, or keywords, and makes every accept decision. Over 114 single-shot generation jobs and 240 closed-loop runs on four commercial models, single-shot prompting met the target in 21.1 to 31.6 percent of cases and the closed loop in 92.5 to 98.8 percent, within two edit rounds on average and at a recall-based fidelity of 0.92 to 0.93; the two models common to both settings show the same effect. Because the controller optimizes the value on which success is scored, the result establishes reproducible control over a declared, computable metric and not validated human difficulty. The transferable practice is to declare the acceptance condition as code, bound the model to local edits, and gate every edit on a content check.

cs.SE↗

arXiv2026

Markov Chain CLTs: Resolving Open Problems

Markov chain central limit theorems (CLTs) and their associated variances are very important for implementing Markov chain Monte Carlo algorithms among other applications. Häggström and Rosenthal (2007) presented various results regarding the equality of different formulae for this variance and also posed seven open problems. We resolve all seven in this paper. For stationary, ergodic and reversible chains, we prove that whenever the normalized partial sums satisfy a $\sqrt{n}$-CLT, the variance limit is finite if the function is square-integrable, otherwise undefined. Moreover, failure of the $\sqrt{n}$-CLT forces the normalized partial sums to be non-tight. We also show that Roberts' holding-probability condition precludes a CLT even without assuming reversibility or square-integrability. Finally, we develop a general principle that expresses Fourier coefficients as the autocovariances of an ergodic nonreversible Markov chain.

math.PR↗

arXiv2026

Agent System Operations: Categorization, Challenges, and Future Directions

As the reasoning capabilities of Large Language Models (LLMs) continue to advance, LLM-based agent systems offer advantages in flexibility and interpretability over traditional systems, garnering increasing attention. However, despite the widespread research interest and industrial application of agent systems, these systems, like their traditional counterparts, frequently encounter anomalies. These anomalies lead to instability and insecurity, hindering their further development. Therefore, a comprehensive and systematic approach to the operation and maintenance of agent systems is urgently needed. Unfortunately, current research on the operations of agent systems is sparse. To address this gap, we have undertaken a survey on agent system operations with the aim of establishing a clear framework for the field, defining the challenges, and facilitating further development. Specifically, this paper begins by systematically defining anomalies within agent systems, categorizing them into intra-agent anomalies and inter-agent anomalies. Next, we introduce a novel and comprehensive operational framework for agent systems, dubbed Agent System Operations (AgentOps). We provide detailed definitions and explanations of its four key stages: monitoring, anomaly detection, root cause localization, and resolution.

cs.MA↗

arXiv2026

Revisiting the Hubble tension with an inverse power-law early dark energy

We propose a new early dark energy with a potential that has an inverse power-law asymptotic tail to alleviate the Hubble tension. Fitting this model to the datasets of CMB+BAO+SN+H0DN, one obtains a 68$\%$ confidence-level value of $H_0\sim 70.20^{+0.60}_{-0.47}$ km s$^{-1}$Mpc$^{-1}$ and a best-fit value of $H_0\sim 70.56$ km s$^{-1}$Mpc$^{-1}$ for index $n=10$. We show that this model can be tested by future data of CMB and matter power spectra. Finally, we briefly discuss fine-tuning problems related to this model.

astro-ph.CO↗

arXiv2026

Interpreting the Hubble tension with a cascade decaying dark matter sector

Hubble tension can be alleviated by altering either early- or late-time $Λ$CDM. With only one of these effects introduced, early dark energy remains the only solution capable of reducing the tension to the $3σ$ level or below. In this work, we instead consider a modification of the dark matter sector that incorporates both the early- and late-time effects, with the goal of achieving the largest possible value of $H_0$ within this framework. As a realization of these two-fold effects, we study a cascade decaying dark matter model. By fitting the model to the latest datasets of Planck CMB+ DESI BAO+Pantheon (+SH0ES), we find that a 68$\%$ CL value of $H_{0}=68.76\pm0.35 (69.05^{+0.31}_{-0.27})$ km s$^{-1}$ Mpc$^{-1}$ with $Δ\rm{AIC}=+22.0(18.4)$, and larger value of $H_0$ can be obtained by adjusting parameter priors but with a cost of significantly increased value of $Δ\rm{AIC}$. These findings revise the earlier results on the tension level in the literature. For completeness, we show that the parameter regions favored by the cosmological datasets are compatible with complementary limits arising from the Big Bang Nucleosynthesis, neutrino flux, and structure formation.

astro-ph.CO↗

arXiv2026

Multi-Source Retrieval and Reasoning for Legal Sentencing Prediction

Legal judgment prediction (LJP) aims to predict judicial outcomes from case facts and typically includes law article, charge, and sentencing prediction. While recent methods perform well on the first two subtasks, legal sentencing prediction (LSP) remains difficult due to its need for fine-grained objective knowledge and flexible subjective reasoning. To address these limitations, we propose $MSR^2$, a framework that integrates multi-source retrieval and reasoning in LLMs with reinforcement learning. $MSR^2$ enables LLMs to perform multi-source retrieval based on reasoning needs and applies a process-level reward to guide intermediate subjective reasoning steps. Experiments on two real-world datasets show that $MSR^2$ improves both accuracy and interpretability in LSP, providing a promising step toward practical legal AI. Our code is available at https://github.com/cjj826/MSR2.

cs.IR↗

arXiv2026

Software Engineering for and with GUI Agent

GUI agents have advanced rapidly, producing a growing body of frameworks, benchmarks, and applications. However, this growth has outpaced the maturity of the field. GUI agents remain technically brittle, incompletely engineered, and insufficiently validated for sustained real-world use. They are evolving into closed-loop software systems. Within these systems, model reasoning is coupled with interface perception, execution feedback, recovery, and human oversight. This evolution calls for a software engineering perspective that remains largely absent from existing research. We address this gap by reviewing 336 GUI-agent papers from January 2018 to April 2026. Five research questions examine the research landscape, architectures, evaluation, software lifecycle concerns, and future opportunities. Our findings show that the field has expanded sharply since 2024, while mobile and web settings remain dominant. Architectures increasingly adopt modular perceive-reason-act loops, but recovery, human escalation, safety enforcement, and auditability remain underdeveloped. This architectural imbalance extends to evaluation. Evaluations are becoming more interactive, but they remain centered on task success and are difficult to compare across protocols. More broadly, existing studies provide limited support for testing beyond benchmarks and for maintaining agents after release. Observability, privacy engineering, and systematic human oversight are also underdeveloped. Together, these findings show that capability improvements alone cannot ensure deployment readiness. Future research should connect dependable execution with lifecycle-centered testing and reproducible evaluation. It should also integrate permission and privacy controls with cost-aware, human-centered governance. This integration is necessary to build dependable, maintainable, secure, and deployable GUI-agent systems.

cs.SE↗

arXiv2026

Chernoff's Density Is Strongly Log-Concave

Let $f$ be the density of the Chernoff random variable $\mathrm{argmax}_{t\in\mathbb{R}}\{W(t)-t^2\}$, where $W$ is a two-sided Brownian motion. This note proves the conjecture of Balabdaoui and Wellner (2014) that $f$ is strongly log-concave. The proof was generated in its entirety by GPT-5.6 Sol.

math.PR↗

arXiv2026

MedRealMM: A Real-World Multimodal Benchmark for Chinese Online Medical Consultation

Large language models (LLMs) are increasingly deployed in online medical consultation, yet existing benchmarks remain poorly aligned with real clinical practice. Many rely on synthetic conversations or patient simulators, omit patient-uploaded medical images, or evaluate open-ended clinical responses using multiple-choice or lexical-overlap metrics that poorly reflect clinical quality. We introduce \textbf{MedRealMM}, a large-scale benchmark for multimodal online medical consultation built from de-identified patient-doctor interactions collected from a nationwide Chinese internet hospital. MedRealMM uses a Multimodal Clinical Challenge Point (MCCP) extraction framework to identify clinically demanding moments in authentic consultation trajectories and converts each into a standardized next-response generation task while preserving the preceding text-image context. Each instance is paired with a case-specific rubric refined by physicians that rewards clinically desirable behaviors and penalizes unsafe, unsupported, or contradictory responses. The current release contains 5,620 real-world multimodal cases spanning 64 clinical departments. We evaluate 19 general-purpose and medical-specialized LLMs, including text-only and multimodal systems. Our results show that image information is critical for reliable clinical performance and that current frontier models remain below the online physician response. Although some frontier models satisfy as many or more positive clinical criteria than physicians, they trigger more negative criteria, indicating that safety-sensitive error avoidance remains a central bottleneck. MedRealMM offers a realistic and reproducible benchmark for evaluating multimodal medical reasoning in real-world online consultation. The dataset will be publicly available on Hugging Face at https://huggingface.co/datasets/jdh-algo/MedRealMM.

cs.AI↗

arXiv2026

Don't Predict, Prioritize: Rethinking GPU Reliability Assessment

The reliability of Graphics Processing Units (GPUs) is a criticalbottleneck for modern large-scale AI infrastructure, where a sin-gle node failure can disrupt synchronous training jobs and causesignificant financial losses. While predictive maintenance is widelyused in other hardware domains, we demonstrate that accuratelypredicting the exact timing of GPU failures is inherently difficult.Through an in-depth analysis of telemetry data from a productioncluster, we find that major GPU failures, including Double Bit Er-rors (DBEs) and GPU Lost events, exhibit strong stochasticity andlow signal-to-noise ratios in time-series telemetry, which makesconventional time-based prediction ineffective. This insight motivates a paradigm shift: instead of attempting topredict the absolute timing of a failure, we propose a more robustapproach focused on ranking nodes by their relative failure risk. Wepropose HeaRank (Health Rank), a Learning-to-Rank (LTR) frame-work that leverages stable historical failure patterns to computea global risk ranking of GPU nodes. Evaluated on a production-scale cluster with thousands of GPUs, HeaRank achieves an AUCof 0.83, significantly outperforming both heuristic baselines andstate-of-the-art ranking algorithms. In online deployment, HeaRanksuccessfully captures 64% of future failures within the top 5% ofranked nodes, compared to only 21% by the incumbent productionsystem. These results suggest that relative risk ranking can serveas a robust alternative in environments where absolute failure pre-diction is inherently limited. Our work highlights the importanceof risk-aware scheduling and proactive resource management inmodern GPU clusters.

cs.DC↗

arXiv2026

Mask to Concept: Auto-Promptable SAM3 via Efficient Test-Time Concept Embedding Search for Few-Shot Annotation

Transforming foundation segmentation models from human-prompted tools into auto-promptable annotators is critical for scalable medical data annotation. Current methods commonly depend on external feature matchers or auxiliary networks to automate geometric prompting, but introducing architectural overhead and limiting performance scalability. Although SAM3 natively supports concept segmentation via reusable text prompts, its direct use in medical imaging is hindered by a lack of fine-grained clinical knowledge and the ambiguity of human-written descriptions. In this work, we propose Mask to Concept (M2C), an efficient framework that adapts SAM3 for medical few-shot annotation without external modules, parameter retraining, or manual text engineering. Using only a few labeled images, M2C enables SAM3 to automatically search for transferable visual concepts entirely within its frozen architecture: it initializes a learnable concept embedding, uses it to prompt segmentation, and updates the embedding by gradients of minimizing the concept segmentation error. We further introduce a Hybrid Uncertainty Estimation (HUE) module that calculates the prediction entropy and maps concept predictions back to the box prompts, measuring concept-geometry prompting inconsistency. Highly uncertain samples are flagged actively for human correction, and the corrected masks are then fed back to M2C to continuously search for more precise concept embeddings, forming a self-enhancing annotation loop with minimal expert effort. Experiments on medical segmentation benchmarks show that our method achieves SOTA few-shot segmentation performance and outstanding annotation efficiency, offering a practical and efficient pathway toward scalable medical image labeling. Codes are at https://github.com/Huster-Hq/M2C.

cs.CV↗

arXiv2026

CoRe: A Continuously Reward-Finetuned LLM Query Rewriter for Multi-Stage Context-Aware Relevance in Web-Scale Video Search

LLM-based query rewriters in production face a tension: the training reward must reflect how the rewrite is consumed by the production ranker, yet the training procedure must be cheap enough to support continuous redeployment as data drifts. We present CoRe (Context Relevance), such a system, redeployed weekly for over five months in a major short-video search engine. Our reward uses the deployed multimodal relevance model as its source and a multiplicative ratio form mirroring the production fusion algebra, closing the simulation-production gap that offline reward proxies leave open. A semi-online Mixed Preference Optimization loop makes this reward affordable at multi-million-instance weekly scale: a DPO-style pairwise objective restricts the gradient pass to a small top-k/bottom-k subset of sampled trajectories, and a phase structure reduces trainer/inference-server parameter syncs from per-step to per-phase. An automated promotion gate over reward-like and stability metrics detected and recovered from a real reward-hacking incident in production. Rewriter output is consumed as parallel relevance signals at recall, rawrank, and finerank without displacing the original signals, bounding rewriter-failure blast radius. Online A/B from two sequential production launches, first deploying the rewriter at finerank, then extending consumption to recall and rawrank, delivers statistically significant reductions in change-query rate on rewrite-impacted queries, with all headline relevance and engagement metrics moving in the expected direction.

cs.IR↗

arXiv2026

Intelligent Domain Adaptation for Power System Transient Stability Assessment Under Varying Operating Scenarios

While deep learning-based transient stability assessment (TSA) approaches have exhibited great potential in power system stability monitoring, they are prone to undergo performance degradation in practical contexts with frequent variations of operating conditions. To address this issue, this work develops an adaptive TSA framework via domain adaptation-enabled deep transfer learning. First, for the sake of capturing the primary transient stability characteristics, a robust metric, i.e., heterogeneous hybrid distribution metric (HHDM), is designed through mathematical means to effectively handle multi-scale Gaussian and long-tail distributions of transient responsive data and to precisely quantify the intrinsic distributional discrepancies between the source and target domains corresponding to different operating scenarios. With the help of the HHDM, a Bayesian theory-based dual-distribution domain adaptation method is constructed, aligning not only marginal probability distributions between domains but also the distributions of sub-domain categories. Such alignments enable fine-grained transient stability feature transfer, helping significantly improve the adaptability of a well-trained TSA model to target domains. Furthermore, a multilayer sparse regularization algorithm is introduced to mitigate feature volatility caused by variations in operating scenarios, thereby enhancing the model's generalization in the presence of unforeseen scenarios. Numerical tests on three test systems illustrate that, compared with conventional methods, the proposed framework improves online TSA accuracy by 0.5% to 5% in a cost-effective manner, with the learning cost for TSA model update largely reduced.

eess.SY↗