arXiv subjects

Fan Wu

Publications and source records attributed to Fan Wu.

At least 109 records · Page 6Linked to original sources

arXiv2026

Can LLMs Clean Up Your Mess? A Survey of Application-Ready Data Preparation with LLMs

Data preparation aims to denoise raw datasets, uncover cross-dataset relationships, and extract valuable insights from them, which is essential for a wide range of data-centric applications. Driven by (i) rising demands for application-ready data (e.g., for analytics, visualization, decision-making), (ii) increasingly powerful LLM techniques, and (iii) the emergence of infrastructures that facilitate flexible agent construction (e.g., using Databricks Unity Catalog), LLM-enhanced methods are rapidly becoming a transformative and potentially dominant paradigm for data preparation. By investigating hundreds of recent literature works, this paper presents a systematic review of this evolving landscape, focusing on the use of LLM techniques to prepare data for diverse downstream tasks. First, we characterize the fundamental paradigm shift, from rule-based, model-specific pipelines to prompt-driven, context-aware, and agentic preparation workflows. Next, we introduce a task-centric taxonomy that organizes the field into three major tasks: data cleaning (e.g., standardization, error processing, imputation), data integration (e.g., entity matching, schema matching), and data enrichment (e.g., data annotation, profiling). For each task, we survey representative techniques, and highlight their respective strengths (e.g., improved generalization, semantic understanding) and limitations (e.g., the prohibitive cost of scaling LLMs, persistent hallucinations even in advanced agents, the mismatch between advanced methods and weak evaluation). Moreover, we analyze commonly used datasets and evaluation metrics (the empirical part). Finally, we discuss open research challenges and outline a forward-looking roadmap that emphasizes scalable LLM-data systems, principled designs for reliable agentic workflows, and robust evaluation protocols.

cs.DB↗

arXiv2026

VIAFormer: Voxel-Image Alignment Transformer for High-Fidelity Voxel Refinement

We propose VIAFormer, a Voxel-Image Alignment Transformer model designed for Multi-view Conditioned Voxel Refinement--the task of repairing incomplete noisy voxels using calibrated multi-view images as guidance. Its effectiveness stems from a synergistic design: an Image Index that provides explicit 3D spatial grounding for 2D image tokens, a Correctional Flow objective that learns a direct voxel-refinement trajectory, and a Hybrid Stream Transformer that enables robust cross-modal fusion. Experiments show that VIAFormer establishes a new state of the art in correcting both severe synthetic corruptions and realistic artifacts on the voxel shape obtained from powerful Vision Foundation Models. Beyond benchmarking, we demonstrate VIAFormer as a practical and reliable bridge in real-world 3D creation pipelines, paving the way for voxel-based methods to thrive in large-model, big-data wave.

cs.CV↗

arXiv2026

TacUMI: A Multi-Modal Universal Manipulation Interface for Contact-Rich Tasks

Task decomposition is critical for understanding and learning complex long-horizon manipulation tasks. Especially for tasks involving rich physical interactions, relying solely on visual observations and robot proprioceptive information often fails to reveal the underlying event transitions. This raises the requirement for efficient collection of high-quality multi-modal data as well as robust segmentation method to decompose demonstrations into meaningful modules. Building on the idea of the handheld demonstration device Universal Manipulation Interface (UMI), we introduce TacUMI, a multi-modal data collection system that integrates additionally ViTac sensors, force-torque sensor, and pose tracker into a compact, robot-compatible gripper design, which enables synchronized acquisition of all these modalities during human demonstrations. We then propose a multi-modal segmentation framework that leverages temporal models to detect semantically meaningful event boundaries in sequential manipulations. Evaluation on a challenging cable mounting task shows more than 90 percent segmentation accuracy and highlights a remarkable improvement with more modalities, which validates that TacUMI establishes a practical foundation for both scalable collection and segmentation of multi-modal demonstrations in contact-rich tasks.

cs.RO↗

arXiv2026

Nonadiabatic transitions in non-Hermitian $\mathcal{PT}$-symmetric two-level systems

We systematically characterize the dynamical evolution of time-parity (PT )-symmetric two-level systems with spin-dependent dissipations. If the control parameters of the gap are linearly tuned with time, the dynamical evolution can be characterized with parabolic cylinder equations which can be analytically solved. We find that the asymptotic behaviors of particle probability on the two levels show initial-state-independent redistribution in the slow-tuning-speed limit as long as the system is nonadiabatically driven across exceptional points. Equal distributions appear when the nondissipative Hamiltonian shows gap closing. So long as the nondissipative Hamiltonian displays level anticrossing, the final distribution becomes unbalanced. The ratios between the occupation probabilities are given analytically. These results are confirmed with numerical simulations. The predicted equal distribution phenomenon may be used to identify the closing of the energy gap from anti-crossing between two energy bands.

quant-ph↗

arXiv2026

Chiral Dynamics Near Intra- and Inter-Band Exceptional Points under Dissipative Spin-Orbital-Angular-Momentum Coupling

We study the parametric chiral dynamics of atoms under dissipative spin-orbital-angular-momentum coupling (SOAMC). With atoms confined in the ring-shaped potential of the Laguerre-Gaussian Raman beams, the SOAMC not only couples the atomic center-of-mass angular momentum to the hyperfine spins, but also mixes different bands in the radial direction. This gives rise to a series of exceptional points of two types, the intra-band and the inter-band. Leveraging the topology of the spectral Riemann surface close to these exceptional points, we demonstrate the path-dependent chiral transfer of atoms to the higher-lying bands, by evolving the system along closed loops in the parameter space. Specifically, we illustrate two distinct scenarios, characterized by different mechanisms, where the atoms can be transferred to designated SOAMC-dressed bands. Our work demonstrates the rich exceptional structure in atom gases under dissipative SOAMC, and offers a novel route toward populating higher bands.

cond-mat.quant-gas↗

arXiv2025

Critical quantum metrology robust against dissipation and non-adiabaticity

Critical systems near quantum phase transitions were predicted to be useful for improvement of metrological precision, thanks to their ultra-sensitive response to a tiny variation of the control Hamiltonian. Despite the promising perspective, realization of criticality-enhanced quantum metrology is an experimentally challenging task, mainly owing to the extremely long time needed to encode the signal to some physical quantity of a critical system. We here circumvent this problem by making use of the critical behaviors in the Jaynes-Cummings model, comprising a single qubit and a photonic resonator, to which the signal field is coupled. The information about the field amplitude is encoded in the qubit's excitation number in the dark state, which displays a divergent changing rate at the critical point. The most remarkable feature of this critical sensor is that the performance is insensitive to the leakage to bright eigenstates, caused by decoherence and non-adiabatic effects. We demonstrate such a metrological protocol in a superconducting circuit, where an Xmon qubit, interacting with a resonator, is used as a probe for estimating the amplitude of a microwave field coupled to the resonator. The measured quantum Fisher information exhibits a critical quantum enhancement, confirming the potential of this system for quantum metrology.

quant-ph↗

arXiv2025

UPETrack: Unidirectional Position Estimation for Tracking Occluded Deformable Linear Objects

Real-time state tracking of Deformable Linear Objects (DLOs) is critical for enabling robotic manipulation of DLOs in industrial assembly, medical procedures, and daily-life applications. However, the high-dimensional configuration space, nonlinear dynamics, and frequent partial occlusions present fundamental barriers to robust real-time DLO tracking. To address these limitations, this study introduces UPETrack, a geometry-driven framework based on Unidirectional Position Estimation (UPE), which facilitates tracking without the requirement for physical modeling, virtual simulation, or visual markers. The framework operates in two phases: (1) visible segment tracking is based on a Gaussian Mixture Model (GMM) fitted via the Expectation Maximization (EM) algorithm, and (2) occlusion region prediction employing UPE algorithm we proposed. UPE leverages the geometric continuity inherent in DLO shapes and their temporal evolution patterns to derive a closed-form positional estimator through three principal mechanisms: (i) local linear combination displacement term, (ii) proximal linear constraint term, and (iii) historical curvature term. This analytical formulation allows efficient and stable estimation of occluded nodes through explicit linear combinations of geometric components, eliminating the need for additional iterative optimization. Experimental results demonstrate that UPETrack surpasses two state-of-the-art tracking algorithms, including TrackDLO and CDCPD2, in both positioning accuracy and computational efficiency.

cs.RO↗

arXiv2025

Exploring the topology induced by non-Markovian Liouvillian exceptional points

Non-Hermitian (NH) systems can display exotic topological phenomena without Hermitian counterparts, enabled by exceptional points (EPs). So far, investigations of NH topology have been restricted to EPs of the NH Hamiltonian, which governs the system dynamics conditional upon no quantum jumps occurring. The Liouvillian superoperator, which combines the effects of quantum jumps with NH Hamiltonian dynamics, possesses EPs (LEPs) that are significantly different from those of the corresponding NH Hamiltonian. We here study the topological features of the LEPs in the system consisting of a qubit coupled to a non-Markovian reservoir. We find that two distinct winding numbers can be simultaneously produced by executing a single closed path encircling the twofold LEP2, formed by two coinciding LEP2s, each involving a pair of coalescing eigenvectors of the extended Liouvillian superoperator. We experimentally demonstrate this purely non-Markovian phenomenon with a circuit, where a superconducting qubit is coupled to a decaying resonator which acts as a reservoir with memory effects. The results push the exploration of exceptional topology from the Markovian to non-Markovian regime.

quant-ph↗

arXiv2025

IC-World: In-Context Generation for Shared World Modeling

Video-based world models have recently garnered increasing attention for their ability to synthesize diverse and dynamic visual environments. In this paper, we focus on shared world modeling, where a model generates multiple videos from a set of input images, each representing the same underlying world in different camera poses. We propose IC-World, a novel generation framework, enabling parallel generation for all input images via activating the inherent in-context generation capability of large video models. We further finetune IC-World via reinforcement learning, Group Relative Policy Optimization, together with two proposed novel reward models to enforce scene-level geometry consistency and object-level motion consistency among the set of generated videos. Extensive experiments demonstrate that IC-World substantially outperforms state-of-the-art methods in both geometry and motion consistency. To the best of our knowledge, this is the first work to systematically explore the shared world modeling problem with video-based world models.

cs.CV↗

arXiv2025

PhyCustom: Towards Realistic Physical Customization in Text-to-Image Generation

Recent diffusion-based text-to-image customization methods have achieved significant success in understanding concrete concepts to control generation processes, such as styles and shapes. However, few efforts dive into the realistic yet challenging customization of physical concepts. The core limitation of current methods arises from the absence of explicitly introducing physical knowledge during training. Even when physics-related words appear in the input text prompts, our experiments consistently demonstrate that these methods fail to accurately reflect the corresponding physical properties in the generated results. In this paper, we propose PhyCustom, a fine-tuning framework comprising two novel regularization losses to activate diffusion model to perform physical customization. Specifically, the proposed isometric loss aims at activating diffusion models to learn physical concepts while decouple loss helps to eliminate the mixture learning of independent concepts. Experiments are conducted on a diverse dataset and our benchmark results demonstrate that PhyCustom outperforms previous state-of-the-art and popular methods in terms of physical customization quantitatively and qualitatively.

cs.CV↗

arXiv2025

Crystal Growth and Physical Properties of Orthorhombic Kagome Lattice Magnets $R$Fe$_6$Ge$_6$ ($R$=Y, Tb, Dy)

Kagome magnets represent a promising class of materials that exhibit intriguing electronic and magnetic properties, and they have recently garnered significant attention. While most kagome-lattice compounds are hexagonal, we report here single-crystal growth and physical property measurements of $R$Fe$_6$Ge$_6$ ($R$ = Y, Dy, Tb) compounds, which crystallize in an orthorhombic structure. The structure can be derived from a hexagonal prototype $R$Fe$_3$Ge$_2$ by replacing every other $R$ atom with a covalent Ge$_2$ dimer. Ordering of these dimers renders the structure orthorhombic, slightly distorts the kagome net, and makes the three Fe sites formally inequivalent. The iron and rare-earth sublattices order independently. Fe moments order above 400 K, forming ferromagnetic kagome planes stacked antiferromagnetically, while rare-earth moments order below 9 K. TbFe$_6$Ge$_6$ exhibits a single magnetic ordering transition associated with the Tb atoms, whereas DyFe$_6$Ge$_6$ shows two distinct magnetic phase transitions, strongly influenced by crystal electric field effects on the Dy$^{3+}$ ions. Density functional theory (DFT) calculations indicate that the ferromagnetic ordering of the Fe planes is driven by a high density of states at the Fermi energy. They also reveal three dramatically different structural energy scales: $R$ and Ge$_2$ form alternating 1D chains perpendicular to the kagome planes, and violating this alternation incurs a large energy cost. Aligning these chains is less costly, and achieving a two-dimensional order of anti-aligned chains requires very little energy. These compounds represent a unique class of materials, offering new opportunities to investigate the interplay between the distinct crystal lattice geometry and the underlying electronic and magnetic properties.

cond-mat.mtrl-sci↗

arXiv2025

TTF: A Trapezoidal Temporal Fusion Framework for LTV Forecasting in Douyin

In the user growth scenario, Internet companies invest heavily in paid acquisition channels to acquire new users. But sustainable growth depends on acquired users' generating lifetime value (LTV) exceeding customer acquisition cost (CAC). In order to maximize LTV/CAC ratio, it is crucial to predict channel-level LTV in an early stage for further optimization of budget allocation. The LTV forecasting problem is significantly different from traditional time series forecasting problems, and there are three main challenges. Firstly, it is an unaligned multi-time series forecasting problem that each channel has a number of LTV series of different activation dates. Secondly, to predict in the early stage, it faces the imbalanced short-input long-output (SILO) challenge. Moreover, compared with the commonly used time series datasets, the real LTV series are volatile and non-stationary, with more frequent fluctuations and higher variance. In this work, we propose a novel framework called Trapezoidal Temporal Fusion (TTF) to address the above challenges. We introduce a trapezoidal multi-time series module to deal with data unalignment and SILO challenges, and output accurate predictions with a multi-tower structure called MT-FusionNet. The framework has been deployed to the online system for Douyin. Compared to the previously deployed online model, MAPEp decreased by 4.3%, and MAPEa decreased by 3.2%, where MAPEp denotes the point-wise MAPE of the LTV curve and MAPEa denotes the MAPE of the aggregated LTV.

cs.LG↗

arXiv2025

EcoAgent: An Efficient Device-Cloud Collaborative Multi-Agent Framework for Mobile Automation

To tackle increasingly complex tasks, recent research on mobile agents has shifted towards multi-agent collaboration. Current mobile multi-agent systems are primarily deployed in the cloud, leading to high latency and operational costs. A straightforward idea is to deploy a device-cloud collaborative multi-agent system, which is nontrivial, as directly extending existing systems introduces new challenges: (1) reliance on cloud-side verification requires uploading mobile screenshots, compromising user privacy; and (2) open-loop cooperation lacking device-to-cloud feedback, underutilizing device resources and increasing latency. To overcome these limitations, we propose EcoAgent, a closed-loop device-cloud collaborative multi-agent framework designed for privacy-aware, efficient, and responsive mobile automation. EcoAgent integrates a novel reasoning approach, Dual-ReACT, into the cloud-based Planning Agent, fully exploiting cloud reasoning to compensate for limited on-device capacity, thereby enabling device-side verification and lightweight feedback. Furthermore, the device-based Observation Agent leverages a Pre-understanding Module to summarize screen content into concise textual descriptions, significantly reducing token usage and device-cloud communication overhead while preserving privacy. Experiments on AndroidWorld demonstrate that EcoAgent matches the task success rates of fully cloud-based agents, while reducing resource consumption and response latency. Our project is available here: https://github.com/Yi-Biao/EcoAgent.

cs.AI↗

arXiv2025

An exceptional surface and its topology

Non-Hermitian (NH) systems can display exceptional topological defects without Hermitian counterparts, exemplified by exceptional rings in NH two-dimensional systems. However, exceptional topological features associated with higher-dimension topological defects remain unexplored yet. We here investigate the topology for the singularities in an NH three-dimensional system. We find that the three-order singularities in the parameter space form an exceptional surface (ES), on which all the three eigenstates and eigenenergies coalesce. Such an ES corresponds to a two-dimensional extension of a point-like synthetic tensor monopole. We quantify its topology with the Dixmier-Douady invariant, which measures the quantized flux associated with the synthetic tensor field. We further propose an experimentally feasible scheme for engineering such an NH model. Our results pave the way for investigations of exceptional topology associated with topological defects with more than one dimension.

quant-ph↗

arXiv2025

Learning Temporal 3D Semantic Scene Completion via Optical Flow Guidance

3D Semantic Scene Completion (SSC) provides comprehensive scene geometry and semantics for autonomous driving perception, which is crucial for enabling accurate and reliable decision-making. However, existing SSC methods are limited to capturing sparse information from the current frame or naively stacking multi-frame temporal features, thereby failing to acquire effective scene context. These approaches ignore critical motion dynamics and struggle to achieve temporal consistency. To address the above challenges, we propose a novel temporal SSC method FlowScene: Learning Temporal 3D Semantic Scene Completion via Optical Flow Guidance. By leveraging optical flow, FlowScene can integrate motion, different viewpoints, occlusions, and other contextual cues, thereby significantly improving the accuracy of 3D scene completion. Specifically, our framework introduces two key components: (1) a Flow-Guided Temporal Aggregation module that aligns and aggregates temporal features using optical flow, capturing motion-aware context and deformable structures; and (2) an Occlusion-Guided Voxel Refinement module that injects occlusion masks and temporally aggregated features into 3D voxel space, adaptively refining voxel representations for explicit geometric modeling. Experimental results demonstrate that FlowScene achieves state-of-the-art performance on the SemanticKITTI and SSCBench-KITTI-360 benchmarks.

cs.CV↗

arXiv2025

LLM/Agent-as-Data-Analyst: A Survey

Large language models (LLMs) and agent techniques have brought a fundamental shift in the functionality and development paradigm of data analysis tasks (a.k.a LLM/Agent-as-Data-Analyst), demonstrating substantial impact across both academia and industry. In comparison with traditional rule or small-model based approaches, (agentic) LLMs enable complex data understanding, natural language interfaces, semantic analysis functions, and autonomous pipeline orchestration. From a modality perspective, we review LLM-based techniques for (i) structured data (e.g., NL2SQL, NL2GQL, ModelQA), (ii) semi-structured data (e.g., markup languages understanding, semi-structured table question answering), (iii) unstructured data (e.g., chart understanding, text/image document understanding), and (iv) heterogeneous data (e.g., data retrieval and modality alignment in data lakes). The technical evolution further distills four key design goals for intelligent data analysis agents, namely semantic-aware design, autonomous pipelines, tool-augmented workflows, and support for open-world tasks. Finally, we outline the remaining challenges and propose several insights and practical directions for advancing LLM/Agent-powered data analysis.

cs.AI↗

arXiv2025

Dynamically generating superflow in a bosonic ring via phase imprinting

Phase imprinting enables the dynamic generation of superflow in bosonic atoms, effectively overcoming traditional limitations such as vortex number constraints and heating effects. However, the mechanisms underlying superflow formation remain insufficiently understood. In this work, we reveal these mechanisms by studying the time evolution of the transferred total angular momentum and the quantized current throughout the phase imprinting process, achieved through numerically solving the time-dependent Schrödinger and Gross-Pitaevskii equations. We demonstrate that the Bose gas dynamically acquires angular momentum through the density depletion induced by the phase imprinting potential, whereas quantized currents emerge from azimuthal phase slips accompanied by complete density depletions. Regarding the impact of system parameters, such as interactions, we find that interactions hinder superflow formation, as the azimuthal density distribution becomes less susceptible to the phase imprinting potential. Our findings offer microscopic insights into the dynamic development of superflow during the phase imprinting process and provide valuable guidance for ongoing experimental efforts.

cond-mat.quant-gas↗

arXiv2025

RAGRouter: Learning to Route Queries to Multiple Retrieval-Augmented Language Models

Retrieval-Augmented Generation (RAG) significantly improves the performance of Large Language Models (LLMs) on knowledge-intensive tasks. However, varying response quality across LLMs under RAG necessitates intelligent routing mechanisms, which select the most suitable model for each query from multiple retrieval-augmented LLMs via a dedicated router model. We observe that external documents dynamically affect LLMs' ability to answer queries, while existing routing methods, which rely on static parametric knowledge representations, exhibit suboptimal performance in RAG scenarios. To address this, we formally define the new retrieval-augmented LLM routing problem, incorporating the influence of retrieved documents into the routing framework. We propose RAGRouter, a RAG-aware routing design, which leverages document embeddings and RAG capability embeddings with contrastive learning to capture knowledge representation shifts and enable informed routing decisions. Extensive experiments on diverse knowledge-intensive tasks and retrieval settings, covering open and closed-source LLMs, show that RAGRouter outperforms the best individual LLM and existing routing methods. With an extended score-threshold-based mechanism, it also achieves strong performance-efficiency trade-offs under low-latency constraints. The code and data are available at https://github.com/OwwO99/RAGRouter.

cs.CL↗