arXiv subjects

Lu Han

Publications and source records attributed to Lu Han.

At least 19 recordsLinked to original sources

arXiv2026

Moiré reconstruction at a twisted SrTiO3 interface

Twisted oxide double layers have emerged as a promising platform for exploring moiré physics, including polar topologies and emergent electronic phases. However, the intrinsic complexity of oxide surfaces, such as contamination, roughness, and the presence of two distinct terminations (AO and BO2), can complicate the identification of buried interfacial structures. Here, using atomically flat SrTiO3 membranes with a predominantly SrO-TiO2 interfacial configuration, we directly visualize a moiré-periodic structural reconstruction at the buried interface that is distinct from the geometric moiré of rigid stacking. The reconstructed interface hosts head-to-head-like polar distortions, moiré-modulated Ti electronic structure, and enhanced microwave admittance in otherwise paraelectric and insulating SrTiO3. Our work demonstrates that, with a well-defined termination, moiré reconstruction could serve as a route to emergent electrical behavior in twisted perovskite oxides.

cond-mat.mtrl-sci↗

arXiv2026

GAGR-Lab: Evaluating Joint Spatial-Geometric and Analytic Function Reasoning

Joint spatial-geometric and analytic function reasoning requires translating a perceived spatial configuration into a symbolic function whose executed curve satisfies geometric constraints. We present GAGR-Lab, a framework for measuring this capability through Cartesian game scenes, explicit function semantics, and authoritative Rust trajectory execution. It distinguishes spatial perception, metric grounding, geometric relations, function interpretation, function construction, and constrained synthesis. We specify four configurable scene-difficulty presets and a prospective 24-cell diagnostic design, while reporting only the subset actually evaluated. A bounded pilot of one hosted model (Llama 3.2 11B Vision Instruct) using two API credentials as execution replicas yields 72 balanced games with 432 attempts, 429 valid provider responses, and no target hits; exploratory ordinary-function prompt variants also fail to hit, while the structured localization interface yields no scoreable outputs. A privileged analytic search control independently succeeds on 600 directional cases from 300 generated scenes, with exact repeatability and 1,200 successful vertical-reflection or translation checks. The framework separates serving reliability, symbolic compliance, and geometric success, and preserves exact model-visible inputs and realized paths. A staged protocol outlines diagnostic calibration, held-out replication, multi-model comparison, and paired robustness tests. The contribution is an operational research framework with an executed pilot and a clearly identified prospective study plan; the full difficulty matrix and comparative model results remain untested.

cs.AI↗

arXiv2026

Rethinking the Information Bottleneck: Structured Decomposition under Label-Induced Partitions

Standard information bottleneck (IB) regularization constrains representations via a single scalar I(Z;X), implicitlytreating all information as homogeneous. However, a single global compression control couples label-relevant structurewith residual within-condition variation, rather than regulating their allocation independently, allowing nuisanceinformation to persist in learned representations. For example, in medical imaging applications, residual variation oftenstems from acquisition conditions, background factors, or subject-specific appearance. This issue becomes particularlypronounced in data-limited settings, where models tend to overfit such variation, hindering generalization. While existingregularization methods can stabilize training, control capacity, or shape representation geometry, they do not explicitlyseparate nuisance-like variation from task-supporting structure. To address this limitation, we revisit IB from a structuredperspective based on a label-induced partition, where condition-level structure and within-condition information playdistinct roles. This leads to a dual-bottleneck formulation: a standard KL term controls global information capacity, while aconditional KL term targets within-condition information. We show that the conditional KL admits an exact decompositioninto a within-condition information term and a prior-mismatch term, explaining its alignment with the design objective.With a simplex-structured conditional prior, the method provides controllable latent geometry and integrates seamlesslyinto existing pipelines. Experiments on classification and segmentation show the clearest gains in low-data classificationand consistent improvements across dense prediction benchmarks.

cs.LG↗

arXiv2026

FAVoR: Measuring and Mitigating Author-Style Homogenization in Federated Personalized Generation

Large language models are increasingly used as personalized writing assistants, but adapting a model across many authors can compromise individual writing style by pulling author-specific signals toward a shared register. Federated parameter-efficient fine-tuning (PEFT) offers a data-local setting for this multi-author adaptation problem: clients keep author text local while sharing compact adapter updates. However, we show that standard aggregation can preserve continuation utility while making different authors' generations less distinguishable in style space, a failure mode we define as author-style homogenization. We evaluate author-style retention with Angular Style Classification Encoder (ASCE)-based diagnostics on our main BlogText benchmark and ASCE-independent external authorship verification. Using this protocol, we find that common federated PEFT baselines can preserve semantic utility while averaging out author-specific signals. To address this homogenization, we instantiate FAVoR (Federated Authorial Voice Retention), an author-style residual mechanism for federated PEFT. FAVoR uses a shared-private adapter design: clients upload shared-adapter updates while retaining author-specific residual corrections locally. Across BlogText and external Mythos-Reddit validation, FAVoR improves author-style retention over standard and personalized federated PEFT baselines. These gains come with small continuation-utility trade-offs and are supported by component ablations, external verification, and cold-start transfer.

cs.CL↗

arXiv2026

Integrated Multivariate Segmentation Tree for Heterogeneous Credit Data Analysis in Small- and Medium-Sized Enterprises

Traditional decision tree models, which rely exclusively on numerical variables, often face challenges in handling high-dimensional data and are limited in their ability to incorporate textual information effectively. To address these limitations, we propose the integrated multivariate segmentation tree (IMST), a comprehensive framework designed to improve credit evaluation for small- and medium-sized enterprises (SMEs) by integrating financial data with textual sources. This method comprises three core stages: (1) transforming textual data into numerical matrices through matrix factorization, (2) selecting salient financial features using Lasso regression, and (3) constructing a multivariate segmentation tree based on either the Gini index or entropy, with weakest-link pruning applied to control model complexity. Experimental results based on a dataset of 1,428 Chinese SMEs demonstrated that IMST achieved an accuracy rate of 88.9%, surpassing both baseline decision trees (87.4%) and conventional models such as support vector machines and neural networks. Furthermore, the proposed model demonstrated superior interpretability and computational efficiency, featuring a more streamlined architecture and improved risk detection capabilities.

cs.LG↗

arXiv2026

POI-Loc: A Fine-Grained POI Localization Benchmark and an Asymmetric Global-to-Local Matching Method

Point-of-interest (POI) localization matches user-provided storefront close-ups to the same shops in wide, geo-tagged vehicle-mounted street views. POIs may change while the surrounding scene stays similar, so scene-level recognition alone cannot establish POI identity. Differences in target scale and capture domains further challenge matching. We introduce POI-Loc, to our knowledge the first benchmark dedicated to this asymmetric, fine-grained POI localization task. Many visual place recognition methods represent each image with a single global vector, which tends to dilute fine-grained features of small storefronts amid background clutter. We propose GLAM (Global-to-Local Asymmetric Matching) to combine global and local evidence. In stage one, a single attention-pooled query probe is matched against compact reference region tokens via learnable soft top-k interaction, with the resulting local similarity fused with global similarity for retrieval. Stage two reuses query region tokens before attention pooling and stored reference tokens for mutual-nearest-neighbor re-ranking. GLAM surpasses both global and two-stage baselines on Recall@1/5/10 and mAP, with about $5\times$ smaller re-ranking features and $280\times$ lower per-pair matching cost than FoL. The benchmark and code will be released at https://github.com/roadhan/glam.

cs.CV↗

arXiv2026

Benchmarking Tabular Foundation Models as Surrogates in Expensive Evolutionary Optimization

Surrogate-assisted evolutionary algorithms (SAEAs) are effective methods for solving expensive optimization problems (EOPs), where surrogate models replace most expensive evaluations and critically influence the final optimization results. In recent years, tabular foundation models have advanced rapidly, and the Tabular Prior-data Fitted Network (TabPFN) has been adopted as a surrogate model for EOPs due to its strong predictive capability, demonstrating promising performance. Motivated by its potential as a surrogate model in SAEAs, this work conducts a comprehensive study that combines extensive experiments with in-depth theoretical analysis to investigate the effectiveness of TabPFN. Specifically, we perform experiments across both offline and online SAEA settings, covering diverse problem scenarios such as single-objective, multi-objective, constrained, combinatorial, mixed-variable, and engineering optimization problems. In addition, we further analyze the advantages and limitations of TabPFN within SAEAs and provide practical guidelines for its application in different optimization settings. Results show that the effectiveness of TabPFN is highly problem dependent, and it cannot replace conventional surrogates universally. Overall, TabPFN should be adopted selectively according to data availability, landscape complexity, search space characteristics, and its role within the algorithm. Customized model management strategies and role-specific algorithm design are necessary to fully exploit its advantages and avoid its pitfalls.

cs.NE↗

arXiv2026

Quasi-Sinusoidal Single Diamond Structure in Royal Jewel Butterfly: An Angle-Independent Photonic Structure

Structural colouration with narrow spectral photonic bandwidth and high reflectivity is of critical importance for modern optical applications, including displays, laser systems, and optical sensing, etc. Achieving such angle independent colouration typically relies on polycrystalline or inherent structural disorder. However, balancing angular uniformity with high brightness and strong colour contrast remains challenging. Herein, we uncover the structural origin of the spectacular bright, angle-independent blue colouration of Hypochrysops polycletus, a sapphire-like Royal Jewel butterfly. Three-dimensional (3D) electron microscopy reveals that the dorsal wing scale has a single diamond structure, a 3D photonic crystal previously documented only in beetles and weevils. The crystal domains form an extraordinary quasi sinusoidal surface geometry with a distinct template morphology-guided arrangement. Unlike typically thicker biophotonic structures that support multiple high symmetry stopbands, this design contains only 3-4 unit cells in the propagation direction. Its optical response is dominated by the fundamental stopband, with two dominant scattering mechanisms: specular reflection at the {111} inclined sidewalls of the hierarchical structure, and funnelling into localised quasi-normal modes enabled by a strongly anisotropic Bloch transport. By mimicking these features with two-photon polymerisation, we artificially reproduced the optical response in the infrared region. The study opens a pathway towards bioinspired brilliant diffuse colouration and angle-robust photonic devices.

physics.optics↗

arXiv2026

Complete Structural Determination of Mesostructural Dodecagonal Quasicrystalline Particles

Quasicrystals have revolutionized our understanding of order in solids by demonstrating exotic structural and physicochemical properties with diverse potential applications. Despite the development of various theoretical models and experimental techniques to describe quasicrystal structures, the precise determination of local three dimensional (3D) arrangements of constituent atoms, or of secondary building units such as clusters or micelles, remains elusive. This challenge is particularly acute in self assembled soft matter quasicrystalline systems, where the complex assembly of molecular groups introduces additional defects and structural modulations. Herein, we report the first complete structural determination of self-assembled mesostructural dodecagonal quasicrystalline particles. Employing advanced electron tomography, combined with dedicated structural tracing and processing workflows, the 3D coordinates of all nodal sites were extracted. This approach reveals that the actual structure deviates from the conventionally assumed tetrahedral close packing geometry, exhibiting diverse coordination environments and displacive fluctuations. We identified and quantified rotational intergrowths arising from node exchange, as well as various defects and disorder, with these features discernible only through 3D analysis. Additionally, we propose a simplified two-layer stacking of isomorphic hexagonal model to form dodecagonal quasicrystal. This work advances our understanding of soft-matter dodecagonal quasicrystals and paves the way for detailed structural elucidation of self-assembled systems.

cond-mat.mes-hall↗

arXiv2026

OmniAD: Detect and Understand Industrial Anomaly via Multimodal Reasoning

While anomaly detection has made significant progress, generating detailed analyses that incorporate industrial knowledge remains a challenge. To address this gap, we introduce OmniAD, a novel framework that unifies anomaly detection and understanding for fine-grained analysis. OmniAD is a multimodal reasoner that combines visual and textual reasoning processes. The visual reasoning provides detailed inspection by leveraging Text-as-Mask Encoding to perform anomaly detection through text generation without manually selected thresholds. Following this, Visual Guided Textual Reasoning conducts comprehensive analysis by integrating visual perception. To enhance few-shot generalization, we employ an integrated training strategy that combines supervised fine-tuning (SFT) with reinforcement learning (GRPO), incorporating three sophisticated reward functions. Experimental results demonstrate that OmniAD achieves a performance of 79.1 on the MMAD benchmark, surpassing models such as Qwen2.5-VL-7B and GPT-4o. It also shows strong results across multiple anomaly detection benchmarks. These results highlight the importance of enhancing visual perception for effective reasoning in anomaly understanding. All codes and models will be publicly available.

cs.CV↗

arXiv2026

SepSeq: A Training-Free Framework for Long Numerical Sequence Processing in LLMs

While transformer-based Large Language Models (LLMs) theoretically support massive context windows, they suffer from severe performance degradation when processing long numerical sequences. We attribute this failure to the attention dispersion in the Softmax mechanism, which prevents the model from concentrating attention. To overcome this, we propose Separate Sequence (SepSeq), a training-free, plug-and-play framework to mitigate dispersion by strategically inserting separator tokens. Mechanistically, we demonstrate that separator tokens act as an attention sink, recalibrating attention to focus on local segments while preserving global context. Extensive evaluations on 9 widely-adopted LLMs confirm the effectiveness of our approach: SepSeq yields an average relative accuracy improvement of 35.6% across diverse domains while reducing total inference token consumption by 16.4% on average.

cs.CL↗

arXiv2026

UniCA: Unified Covariate Adaptation for Time Series Foundation Model

Time Series Foundation Models (TSFMs) have achieved remarkable success through large-scale pretraining. However, their design primarily targets real-valued series, limiting their ability to handle general forecasting tasks involving diverse and often heterogeneous covariates -- such as categorical variables and multimodal data (e.g., images, text) -- which are typically task-specific and difficult to leverage during pretraining. To address this gap, we propose Unified Covariate Adaptation (UniCA), a framework to bridge TSFMs with general covariate-aware forecasting. UniCA first performs covariate homogenization to transform heterogeneous covariates into high-level homogeneous series representations and then fuses them via a unified attention-based fusion mechanism. UniCA is compatible and universal for adaptation with both homogeneous and heterogeneous covariates, incorporating extra covariate information while preserving the generalization ability of TSFMs.Extensive experiments on multiple unimodal and multimodal covariate-aware forecasting benchmarks demonstrate the superiority of UniCA, highlighting the promise of covariate-aware TSFM adaptation in real-world forecasting scenarios.Code: https://github.com/hanlu-nju/UniCA.

cs.LG↗

arXiv2026

Rethinking ANN-based Retrieval: Multifaceted Learnable Index for Large-scale Recommendation System

Approximate nearest neighbor (ANN) search is widely used in the retrieval stage of large-scale recommendation systems. In this stage, candidate items are indexed using their learned embedding vectors, and ANN search is executed for each user (or item) query to retrieve a set of relevant items. However, ANN-based retrieval has two key limitations. First, item embeddings and their indices are typically learned in separate stages: indexing is often performed offline after embeddings are trained, which can yield suboptimal retrieval quality-especially for newly created items. Second, although ANN offers sublinear query time, it must still be run for every request, incurring substantial computation cost at industry scale. In this paper, we propose MultiFaceted Learnable Index (MFLI), a scalable, real-time retrieval paradigm that learns multifaceted item embeddings and indices within a unified framework and eliminates ANN search at serving time. Specifically, we construct a multifaceted hierarchical codebook via residual quantization of item embeddings and co-train the codebook with the embeddings. We further introduce an efficient multifaceted indexing structure and mechanisms that support real-time updates. At serving time, the learned hierarchical indices are used directly to identify relevant items, avoiding ANN search altogether. Extensive experiments on real-world data with billions of users show that MFLI improves recall on engagement tasks by up to 11.8\%, cold-content delivery by up to 57.29\%, and semantic relevance by 13.5\% compared with prior state-of-the-art methods. We also deploy MFLI in the system and report online experimental results demonstrating improved engagement, less popularity bias, and higher serving efficiency.

cs.IR↗

arXiv2025

Spin-polarized chiral ZnIn2S4 for targeted solar-driven CO2 reduction to acetic acid

Acetic acid, an important industrial chemical, is a key target product for CO2 reduction due to its dual role in carbon utilization and chemical feedstock supply. Although photocatalytic CO2 reduction (PCCR) can generate acetic acid alongside other multicarbon products, its yield is typically low, limited by competing reactions and inefficient C-C coupling. Herein, we report a chiral mesostructured ZnIn2S4 (CMZI) photocatalyst that achieves a remarkable acetic acid yield of 962 {umol g-1 h-1 with a high selectivity of 97.3 %. This yield is ten times higher than the current highest reported value, while attaining state-of-the-art selectivity10. The remarkable productivity arises from synergistic effect between chiral structure and sulfur (S) sites of CMZI. Chirality-induced spin polarization in CMZI stabilizes the key triplet OCCO intermediate, significantly promoting C-C coupling efficiency. Theoretical calculations reveal that the S sites on {102} crystal facets of ZnIn2S4 exhibit thermodynamic and kinetic preferences for acetic acid formation. This work offers critical insights into catalytic strategies for CO2 reduction toward the efficient and scalable synthesis of various multicarbon products.

cond-mat.mtrl-sci↗

arXiv2025

One-Embedding-Fits-All: Efficient Zero-Shot Time Series Forecasting by a Model Zoo

The proliferation of Time Series Foundation Models (TSFMs) has significantly advanced zero-shot forecasting, enabling predictions for unseen time series without task-specific fine-tuning. Extensive research has confirmed that no single TSFM excels universally, as different models exhibit preferences for distinct temporal patterns. This diversity suggests an opportunity: how to take advantage of the complementary abilities of TSFMs. To this end, we propose ZooCast, which characterizes each model's distinct forecasting strengths. ZooCast can intelligently assemble current TSFMs into a model zoo that dynamically selects optimal models for different forecasting tasks. Our key innovation lies in the One-Embedding-Fits-All paradigm that constructs a unified representation space where each model in the zoo is represented by a single embedding, enabling efficient similarity matching for all tasks. Experiments demonstrate ZooCast's strong performance on the GIFT-Eval zero-shot forecasting benchmark while maintaining the efficiency of a single TSFM. In real-world scenarios with sequential model releases, the framework seamlessly adds new models for progressive accuracy gains with negligible overhead.

cs.LG↗

arXiv2025

Advanced spectral clustering for heterogeneous data in credit risk monitoring systems

Heterogeneous data, which encompass both numerical financial variables and textual records, present substantial challenges for credit monitoring. To address this issue, we propose Advanced Spectral Clustering (ASC), a method that integrates financial and textual similarities through an optimized weight parameter and selects eigenvectors using a novel eigenvalue-silhouette optimization approach. Evaluated on a dataset comprising 1,428 small and medium-sized enterprises (SMEs), ASC achieves a Silhouette score that is 18% higher than that of a single-type data baseline method. Furthermore, the resulting clusters offer actionable insights; for instance, 51% of low-risk firms are found to include the term 'social recruitment' in their textual records. The robustness of ASC is confirmed across multiple clustering algorithms, including k-means, k-medians, and k-medoids, with ΔIntra/Inter < 0.13 and ΔSilhouette Coefficient < 0.02. By bridging spectral clustering theory with heterogeneous data applications, ASC enables the identification of meaningful clusters, such as recruitment-focused SMEs exhibiting a 30% lower default risk, thereby supporting more targeted and effective credit interventions.

cs.LG↗

arXiv2025

Deep Learning-Assisted Fourier Analysis for High-Efficiency Structural Design: A Case Study on Three-Dimensional Photonic Crystals Enumeration

The geometric design of structures with optimized physical and chemical properties is one of the core topics in materials science. However, designing new functional materials is challenging due to the vast number of existing and the possible unknown structures to be enumerated and difficulties in mining the underlying correlations between structures and their properties. Here, we propose a universal method for periodic structural design and property optimization. The key in our approach is a deep-learning assisted inverse Fourier transform, which enables the creation of arbitrary geometries within crystallographic space groups. It effectively explores extensive parameter spaces to identify ideal structures with desired properties. Taking the research of three-dimensional (3D) photonic structures as a case study, this method is capable of modelling numerous structures and identifying their photonic bandgaps in just a few hours. We confirmed the established knowledge that the widest photonic bandgaps exist in network morphologies, among which the single diamond (dia net) reigns supreme. Additionally, this method identified a rarely-known lcs topology with excellent photonic properties, highlighting the infinitely extensible application boundaries of our approach. This work demonstrates the high efficiency and effectiveness of the Fourier-based method, advancing material design and providing insights for next-generation functional materials.

physics.optics↗

arXiv2025

WTFormer: A Wavelet Conformer Network for MIMO Speech Enhancement with Spatial Cues Peservation

Current multi-channel speech enhancement systems mainly adopt single-output architecture, which face significant challenges in preserving spatio-temporal signal integrity during multiple-input multiple-output (MIMO) processing. To address this limitation, we propose a novel neural network, termed WTFormer, for MIMO speech enhancement that leverages the multi-resolution characteristics of wavelet transform and multi-dimensional collaborative attention to effectively capture globally distributed spatial features, while using Conformer for time-frequency modeling. A multi task loss strategy accompanying MUSIC algorithm is further proposed for optimization training to protect spatial information to the greatest extent. Experimental results on the LibriSpeech dataset show that WTFormer can achieve comparable denoising performance to advanced systems while preserving more spatial information with only 0.98M parameters.

eess.AS↗