arXiv subjects

Lu Zhang

Publications and source records attributed to Lu Zhang.

At least 55 records · Page 3Linked to original sources

arXiv2026

A Robust Helmholtz-Decomposition-Based Real Compressed Layer Method for Time-Harmonic Elastic Wave Scattering

Time-harmonic elastic wave scattering involves both compressional (P-) and shear (S-) waves, which propagate with different wavenumbers and polarization characteristics. The naive construction of perfectly matched layer (PML)-type methods based on complex coordinate stretching may lack robustness, or even fail, particularly when the wavenumbers are highly contrasted. The recently developed real compressed layer (RCL) technique build upon real compression transformations and explicit extraction of resulting oscillatory patterns for time-harmonic Helmholtz problems may not work, since the oscillations cannot be explicitly extracted by a single change of variables. This paper intends to bridge this gap by developing a robust RCL method for two-dimensional time-harmonic elastic wave scattering in unbounded domains with compactly supported inhomogeneities. A key observation is that, through the Helmholtz decomposition, the displacement field in the exterior homogeneous region decoupled into P-wave and S-wave and each has a distinctive separation of its oscillatory pattern and decaying behaviours in polar coordinates. We then apply the real compression coordinate transformation in the radial direction to each component. We further propose a coupled displacement-potential RCL formulation that seamlessly integrates the Helmholtz-decomposed wave components with the interior displacement field. We show that, under this framework, the essential oscillations in the layer can be effectively removed. We prove the well-posedness of the resulting coupled problem and establish the exponential convergence of the RCL solution to the original scattering solution in the truncated domain of interest. We discretize the RCL-system using high-order spectral element method and demonstrate the effectiveness and robustness of the proposed method through ample numerical results.

math.NA↗

arXiv2026

Linear and nonlinear optical tuning with Sb$_2$S$_3$-based metasurfaces

Here, we experimentally demonstrate the unique properties of Sb$_2$S$_3$ metasurfaces on monolithic and hybridised Sb$_2$S$_3$-Si platforms. Their advantages are particularly pronounced in the telecommunication spectral range, where both phases remain nearly lossless together with a large and stable refractive index contrast compared to other counterparts. The monolithic metasurface enables transmission modulation depths of up to 92\% and resonance shifts as large as $\sim$150~nm at the telecommunication wavelengths through laser-induced phase transitions from amorphous to $\sim$ 50\% polycrystalline states. Furthermore, we demonstrate that integrating a silicon overlayer introduces high-Q hybrid resonances with enhanced near-field confinement, enabling comparable modulation, experimentally, with nearly half the laser switching power required as compared to the monolithic structure. Beyond linear optical switching, the broad resonance tunability of the Sb$_2$S$_3$ metasurface, together with the strong third-order nonlinearity of these materials, provides a promising platform for tunable nonlinear light generation. By exciting the Sb$_2$S$_3$ metasurfaces with a fixed broadband pump, we experimentally demonstrate tunable third-harmonic generation emission over a $\sim$40~nm spectral range through phase-change transition. Such tunability provides a versatile route for integrating Sb$_2$S$_3$ with intrinsically highly nonlinear materials to enable the combination of large optical tunability and efficient nonlinear light generation. We have proven this concept via our hybrid Sb$_2$S$_3$-Si platform, which shows a threefold enhancement in tunable THG emission arising from the combined nonlinear responses of Sb$_2$S$_3$ and silicon, and benefiting from enhanced electromagnetic field confinement enabled by the metasurface's high-Q resonances.

physics.optics↗

arXiv2026

Versioned Late Materialization for Ultra-Long Sequence Training in Recommendation Systems at Scale

Modern Deep Learning Recommendation Models (DLRMs) follow scaling laws with sequence length, driving the frontier toward ultra-long User Interaction History (UIH). However, the industry-standard "Fat Row" paradigm, which pre-materializes these sequences into every training example, creates a storage and I/O wall where data infrastructure usage exceeds GPU training capacity due to data redundancy that is amplified in multi-tenant environments where models with vastly different sequence length requirements share a union dataset. We present a \emph{versioned late materialization} paradigm that eliminates this redundancy by storing UIH once in a normalized, immutable tier and reconstructing sequences just-in-time during training via lightweight versioned pointers. The system ensures Online-to-Offline (O2O) consistency through a bifurcated protocol that prevents future leakage across both streaming and batch training, while a read-optimized immutable storage layer provides multi-dimensional projection pushdown for heterogeneous model tenants. Disaggregated data preprocessing with pipelined I/O prefetching and data-affinity optimizations masks the latency of training-time sequence reconstruction, keeping training throughput compute-bound by GPUs. Deployed on production DLRMs, the system reduces training data infrastructure resource usage while enabling aggressive sequence length scaling that delivers significant model quality gains, serving as the foundational data infrastructure for modern recommendation model architectures, including HSTU and ULTRA-HSTU.

cs.IR↗

arXiv2026

Multiple critical Froude numbers for the centrifugal effects on heat transport in rotating Rayleigh-Bénard convection

The influence of centrifugal effects in rotating Rayleigh-Benard convection is investigated using direct numerical simulations. We find that the Nusselt number decreases beyond a critical Froude number, Fr_c*. This critical value depends on both the Rayleigh number Ra and the aspect ratio Gamma, following power-law scalings with each parameter. We interpret Fr_c* as the onset of centrifugal effects within the thermal boundary layers. This interpretation is supported by the thickening of the boundary layers and a reduction in the planar heat flux. We compare Fr_c* with two previously proposed critical Froude numbers. The first, Fr_Hu, marks the onset of centrifugal effects in the bulk, as evidenced by changes in local heat flux and radial vortex motion. For Fr_Hu < Fr < Fr_c*, centrifugal effects primarily redistribute heat within the bulk and have little influence on the global heat transfer. The second, Fr_Horn, is based on a global force-balance argument. The similar dependence of Fr_c* and Fr_Horn on the aspect ratio suggests a close connection between the global force balance and the onset of centrifugal effects in the thermal boundary layers. These results demonstrate that centrifugal forcing influences the bulk flow and the thermal boundary layers differently in rotating Rayleigh-Benard convection. While relatively weak centrifugal forcing modifies the bulk dynamics, substantially stronger forcing is required to alter boundary-layer properties and global heat transport.

physics.flu-dyn↗

arXiv2026

DisasterBench: A Multimodal Benchmark for UAV-Based Disaster Response in Complex Environments

When a disaster unfolds, responders must answer not only what is happening, but also why it is happening, what will happen next, and what to do now, often from noisy low-altitude UAV views and under tight on-site compute constraints. However, most existing multimodal benchmarks emphasize perception (e.g., recognition/description), cover limited disaster types, and provide insufficient support for the multi-stage reasoning required in practical emergency response. We introduce DisasterBench, a multi-stage multimodal reasoning benchmark for UAV-Based disaster response in complex environments. DisasterBench spans 14 disaster-related scene types and 9 response-critical tasks across pre-, during-, and post-disaster stages, with fine-grained disaster-task mappings that explicitly test causal attribution, propagation prediction, damage analysis, and decision-oriented reasoning. To enable reasoning on the edge, we further propose DisasterVL, a lightweight multimodal model optimized with a three-stage pipeline combining domain instruction tuning, chain-of-thought-guided multimodal alignment, and reinforcement learning-based policy optimization. Experiments across 21 popular MLLMs show that our 2B-parameter DisasterVL outperforms all evaluated open-source models and substantially narrows the gap to state-of-the-art closed-source models, achieving GPT-4o-comparable reasoning accuracy with superior efficiency. The project page is available at https://github.com/TanmouTT/DisasterBench.

cs.CV↗

arXiv2026

An experimental study on the heat transport in porous media convection

We investigate the heat transport in porous media convection over a wide Rayleigh--Darcy number range of $26.8\leq Ra\leq 2.62\times 10^5$, and a Darcy number range of $6.18\times10^{-7}\leq Da\leq 1.21\times 10^{-5}$. In the experiments, we employ 3D-printed lattice structures as the solid porous matrix and water as the working fluid. Quantitative analyses of the porous medium Nusselt number $Nu_m$ and local temperature statistics reveal that the present system undergoes a transition through five distinct regimes: I. Conduction, II. Convection, III. Oscillation, IV. Transition, V. Classical Rayleigh--Bénard convection. This transitional process bridges the gap between Rayleigh--Darcy-like behaviour and Rayleigh--Bénard-like behaviour in porous media convection. By varying the permeability of the matrix, we further examine the role of the Darcy number $Da$, which turns out to have a profound impact on the transitional processes across different regimes. Flow field measurements reveal that the flow structures within Regime IV and Regime V evolve from several horizontally stacked convection rolls to a single-roll structure, and the pore-scale Reynolds number both exceeds unity in these two regimes. Finally, we report the corresponding phase diagram in the $Ra$-$Da$ space.

physics.flu-dyn↗

arXiv2026

Subcritical transition to turbulence in buoyancy-driven flows with multiple hysteresis loops under quasi-one-dimensional confinement

We present both static and quasi-static direct numerical simulations of Rayleigh-Bénard convection in a quasi-one-dimensional domain, revealing for the first time a clear subcritical transition to turbulence in a buoyancy-driven flow. Within a narrow range of Rayleigh number (Ra), three coexisting flow states are identified: steady convection, oscillatory chaos, and intermittent turbulence. The transitions between these states are accompanied by abrupt jumps in both the Nusselt number (Nu) and Reynolds number (Re), the key global transport quantities in buoyancy-driven flows. Additionally, they exhibit pronounced hysteresis, forming three distinct hysteresis loops in the Nu-Ra plane: normal, reverse, and anomalous loops. More importantly, we show that the steady convection state is linearly stable against infinitesimal perturbations but can transition to intermittent turbulence when subjected to finite-amplitude disturbances, which is a defining hallmark of subcriticality. Thus, contrary to the prevailing view that the transition from convection to turbulence is supercritical, our results demonstrate that buoyancy-driven turbulence can emerge via a subcritical route, paving the way for a unified framework that describes instability mechanisms in both buoyancy-driven and shear-driven flows.

physics.flu-dyn↗

arXiv2026

From Accounting to Coordination: A Virtual Water-Aware Electricity-Computation-Water Nexus Framework for Data Center Dispatch

The expansion of data centers (DCs) drives a sustained increase in electricity demand and associated water withdrawals at generation sites. These withdrawals occur at generation sites and are virtually allocated to demand based on network power flows. Consequently, the actual water footprint of a specific load varies dynamically with generation dispatch and network conditions. Existing approaches typically rely on static statistical accounting to quantify these water footprints. However, such static methods fail to capture how dispatch optimization and workload relocation dynamically affect water withdrawals. As a result, static statistical accounting approaches remain decoupled from the optimization process, rendering them incapable of guiding workload relocation or power dispatch to mitigate water stress. To address this limitation, this paper develops an operational electricity-computation-water (ECW) nexus framework that internalizes virtual water impacts directly into power system dispatch. The framework represents dispatch optimization as a differentiable optimization layer embedded within a deep learning architecture, enabling efficient end-to-end learning of coordination policies while preserving operational feasibility. Combined with fixed-point coordination, the framework enforces consistency between virtual water attribution and physical generation-side withdrawals. Case studies on the IEEE 30-bus and 118-bus test systems demonstrate reliable convergence, exact power-water consistency, and reductions of approximately 3-5% in generation-related freshwater withdrawals under water-constrained conditions.

cs.AI↗

arXiv2026

A Fully Discrete Energy-Based Discontinuous Galerkin Method for Variable-Order Time-Fractional Wave Equations

Variable-order time-fractional wave equations provide a flexible model for wave phenomena with evolving memory effects and anomalous temporal dynamics. Their numerical approximation is challenging because the variable-order fractional derivative generates time-dependent history weights and therefore lacks the standard time-translation-invariant convolution structure of constant-order fractional operators. In this paper, we develop and analyze a fully discrete energy-based discontinuous Galerkin (DG) method for wave equations with a Caputo-type variable-order time-fractional derivative. The equation is reformulated as a reduced first-order-in-time system, discretized in space by an energy-based DG method, and advanced in time using a second-order approximation of the variable-order Caputo derivative at a specially chosen point in each time interval. The main analytical novelty is a cumulative weight-variation estimate for the variable-order memory weights, which requires only that the variable order $α:[0,T] \rightarrow (0,1)$ be Lipschitz continuous. Based on this estimate, we establish energy stability of the fully discrete scheme and derive second-order temporal convergence together with energy-norm spatial error estimates. The analysis gives suboptimal convergence on general affine simplicial or tensor-product meshes and optimal convergence under additional Cartesian and flux assumptions. Numerical experiments in one and two dimensions validate the theoretical findings.

math.NA↗

arXiv2026

Contextualized Code Pretraining for Code Generation

As code generation becomes increasingly central to improving software development efficiency, modern code models are largely trained and evaluated on code with natural-language descriptions. In real projects, developers often implement missing functions under limited project-specific artifacts, while the local call-site context is already available in the surrounding code. This usage context provides actionable cues about expected behavior, but existing models are not explicitly optimized to leverage it reliably, leading to implementations that may not integrate smoothly with surrounding usage in repository settings. In this work, we propose contextualized code pretraining, an invocation-aware framework that integrates calling context into both the training and evaluation of code models. Using static analysis, we automatically extract large-scale caller-callee pairs from real repositories to construct pretraining tasks and benchmarks that condition generation on the calling context. We train CallerGen, the first code models pretrained with invocation-aware objectives spanning multiple sizes, and evaluate them on CallerEval, a new benchmark featuring realistic scenarios. Experiments show that CallerGen outperforms comparable-scale models and remains competitive with larger ones across two benchmarks. Our 220M and 0.5B models achieve 16.58% and 22.81@% pass1, surpassing baselines on CallerEval. These results highlight the importance of calling context in realistic code generation.

cs.SE↗

arXiv2026

SeesawNet: Towards Non-stationary Time Series Forecasting with Balanced Modeling of Common and Specific Dependencies

Instance normalization (IN) is widely used in non-stationary multivariate time series forecasting to reduce distribution shifts and highlight common patterns across samples. However, IN can over-smooth instance-specific structural information that is essential for modeling temporal and cross-channel heterogeneity. While prior methods further suppress distribution discrepancies or attempt to recover temporal specific dependencies, they often ignore a central tension: how to adaptively model common and instance-specific dependency based on each instance's non-stationary structures. To address this dilemma, we propose SeesawNet, a unified architecture that dynamically balances common and instance-specific dependency modeling in both temporal and channel dimensions. At its core is Adaptive Stationary-Nonstationary Attention (ASNA), which captures common dependencies from normalized sequences and specific dependencies from raw sequences, and adaptively fuses them according to instance-level non-stationarity. Built upon ASNA, SeesawNet alternates dedicated temporal and channel relationship modeling to jointly capture long-range and cross-variable dependencies. Extensive experiments on multiple real-world benchmarks demonstrate that SeesawNet consistently outperforms state-of-the-art methods.

cs.LG↗

arXiv2026

Synthetic Control Method with Mixed Frequency Data

Mixed-frequency data, where variables are observed at different temporal resolutions, commonly occur in economic and financial studies. Classical synthetic control methods (SCM) are ill-suited for such data, often necessitating aggregation or prefiltering that may discard valuable information. This paper proposes a novel Mixed-Frequency Synthetic Control Method (MF-SCM) to integrate mixed-frequency data into the synthetic control framework effectively. We develop a flexible estimation procedure to construct synthetic control weights under mixed-frequency settings and establish the theoretical properties of the MF-SCM estimator. Specifically, we first prove that the estimator achieves asymptotic optimality, in the sense that it achieves the lowest possible squared prediction error among all potential treatment effect estimators from averaging outcomes of control units. We then derive the asymptotic distribution of the average treatment effect (ATE) estimator using projection theory and construct confidence intervals for the ATE estimator. The method's effectiveness is demonstrated through numerical simulations and two empirical applications concerning the 2017 Tax Cuts and jobs Act in US and air pollution alerts.

stat.ME↗

arXiv2026

Asymptotic Properties of the Distributional Synthetic Controls

As an alternative to synthetic control, the distributional Synthetic Control (DSC) proposed by Gunsilius (2023) provides estimates for quantile treatment effect and thus enabling researchers to comprehensively understand the impact of interventions in causal inference. But the asymptotic properties of DSC have not been built. In this paper, we first establish the DSC estimator's asymptotic optimality in the essence that the treatment effect estimator given by DSC achieves the lowest possible squared prediction error among all potential estimators from averaging quantiles of control units. We then establish the convergence rate of the DSC weights. A significant aspect of our research is that we find the DSC synthesis forms an optimal weighted average, particularly in situations where it is impractical to perfectly fit the treated unit's quantiles through the weighted average of the control units' quantiles. Simulation results verify our theoretical insights.

econ.EM↗

arXiv2026

Retrieving Any Relevant Moments: Benchmark and Models for Generalized Moment Retrieval

Video Moment Retrieval (VMR) aims to localize temporal segments in videos that correspond to a natural language query, but typically assumes only a single matching moment for each query. This assumption does not always hold in real-world scenarios, where queries may correspond to multiple or no moments. Thus, we formulate Generalized Moment Retrieval (GMR), a unified setting that requires retrieving the complete set of relevant moments or predicting an empty set. To enable systematic study of GMR, we introduce Soccer-GMR, a large-scale benchmark built on challenging soccer videos that reflect general GMR scenarios, with realistic negative and positive queries. The benchmark is constructed via a duration-flexible semi-automated pipeline with human verification, enabling scalable data generation while maintaining high annotation quality. We further design a unified evaluation protocol with complementary metrics tailored for null-set rejection, positive-query localization, and end-to-end GMR performance. Finally, we establish strong baselines across two modeling paradigms: a lightweight plug-and-play GMR adapter for discriminative VMR models, and a GMR-tailored GRPO reward for fine-tuning multimodal large language models (MLLMs). Extensive experiments show consistent gains across all metrics and expose key limitations of current methods, positioning GMR as a more realistic and challenging benchmark for video-language understanding.

cs.CV↗

arXiv2026

When Model Editing Meets Service Evolution: A Knowledge-Update Perspective for Service Recommendation

The rapid evolution of software services poses substantial challenges to the design and implementation of effective recommendation systems. Traditional service recommendation approaches often rely on static representations and historical usage data, which are insufficient for adapting to the dynamic and evolving nature of service ecosystems. Recently, large language models (LLMs) have shown strong potential to overcome these limitations by leveraging rich contextual understanding. However, their practical use faces two major challenges: outdated service facts and invalid or redundant services. To address these issues, we propose EVOREC, an evolution-aware framework for service recommendation that leverages model editing in a locate-then-edit paradigm to incorporate updated service facts without costly retraining efficiently. This allows the model to remain aligned with evolving service ecosystems. To address invalid service issues, we introduce a Finite Automata (FA)-based constrained decoding mechanism with deduplication, which enforces structural and semantic validity while eliminating repeated services. Experiments on real-world service datasets demonstrate that our framework consistently outperforms existing baselines, e.g., achieving an average relative improvement of 25.9% in Recall@5. Moreover, under evolving service scenarios, our approach outperforms model fine-tuning approaches by 22.3%, demonstrating strong adaptability to service evolution and providing a practical solution for service recommendation in dynamic ecosystems

cs.SE↗

arXiv2026

CORP: A Multi-Modal Dataset for Campus-Oriented Roadside Perception Tasks

Numerous roadside perception datasets have been introduced to propel advancements in autonomous driving and intelligent transportation systems research and development. However, it has been observed that the majority of their concentrates is on urban arterial roads, inadvertently overlooking residential areas such as parks and campuses that exhibit entirely distinct characteristics. In light of this gap, we propose CORP, which stands as the first public benchmark dataset tailored for multi-modal roadside perception tasks under campus scenarios. Collected in a university campus, CORP consists of over 205k images plus 102k point clouds captured from 18 cameras and 9 LiDAR sensors. These sensors with different configurations are mounted on roadside utility poles to provide diverse viewpoints within the campus region. The annotations of CORP encompass multi-dimensional information beyond 2D and 3D bounding boxes, providing extra support for 3D seamless tracking and instance segmentation with unique IDs and pixel masks for identifying targets, to enhance the understanding of objects and their behaviors distributed across the campus premises. Unlike other roadside datasets about urban traffic, CORP extends the spectrum to highlight the challenges for multi-modal perception in campuses and other residential areas.

cs.CV↗

arXiv2026

Conditional Evidence Reconstruction and Decomposition for Interpretable Multimodal Diagnosis

Neurobiological and neurodegenerative diseases are inherently multifactorial, arising from coupled influences spanning genetic susceptibility, brain alterations, and environmental and behavioral factors. Multimodal modeling has therefore been increasingly adopted for disease diagnosis by integrating complementary evidence across data sources. However, in both large-scale cohorts and real-world clinical workflows, modality coverage is often incomplete, making many multimodal models brittle when one or more modalities are unavailable. Existing approaches to incomplete multimodal diagnosis typically rely on group-wise or static priors, which may fail to capture subject-specific cross-modal dependencies; moreover, many models provide limited interpretability into which evidence sources drive the final decision. To address these limitations, we propose Conditional Evidence Reconstruction and Decomposition (CERD), a framework for interpretable multimodal diagnosis with incomplete modalities. CERD first reconstructs missing modality representations conditioned on each subject's observed inputs, then decomposes diagnostic evidence into shared cross-modal corroboration and modality-specific cues via logit-level attribution. Experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) demonstrate that CERD outperforms competitive baselines under incomplete-modality settings while producing structured and clinically aligned evidence attributions for trustworthy decision support.

cs.CV↗

arXiv2026

LLM-Enhanced Multi-Agent Reinforcement Learning with Expert Workflow for Real-Time P2P Energy Trading

Real-time peer-to-peer (P2P) electricity markets dynamically adapt to fluctuations in renewable energy and variations in demand, maximizing economic benefits through instantaneous price responses while enhancing grid flexibility. However, scaling expert guidance for massive personalized prosumers poses critical challenges, including diverse decision-making demands and a lack of customized modeling frameworks. This paper proposes an integrated large language model-multi-agent reinforcement learning (LLM-MARL) framework for real-time P2P energy trading to address challenges such as the limited technical capability of prosumers, the lack of expert experience, and security issues of distribution networks. LLMs are introduced as experts to generate personalized strategies, guiding MARL under the centralized training with decentralized execution (CTDE) paradigm through imitation. To handle the scalability issues inherent in large-scale P2P networks, a differential attention-based critic network is introduced to efficiently extract key interaction features and enhance convergence. Experimental results demonstrate that LLM-generated strategies effectively substitute human experts. The proposed imitative expert MARL algorithms achieve significantly lower economic costs and voltage violation rates on test sets compared to baseline algorithms, while maintaining robust stability. This paper provides an effective solution for the real-time decision-making of the P2P electricity market by bridging expert knowledge with agent learning.

cs.MA↗