arXiv subjects

Weixuan Liu

Publications and source records attributed to Weixuan Liu.

7 recordsLinked to original sources

arXiv2026

LexPath: A domain-oriented multi-path framework for legal article retrieval

Legal article retrieval is critical for building traceable and reliable legal AI systems, where conclusions must be grounded in specific legal articles. However, general-purpose retrieval methods rely heavily on lexical or semantic similarity, making it difficult to distinguish legally relevant articles from textually similar but legally inapplicable ones, particularly when they differ in their underlying legal intent. To bridge this gap, we propose LexPath, a domain-oriented multi-path framework comprising a multi-path retrieval module and an intent-aware reranking module. The retrieval module combines two complementary domain-specific paths to collect candidate articles: an IRAC-guided sparse path that expands queries with legally informative keywords, and a structure-guided dense path trained with hard negatives derived from legal hierarchy and citation relations. The reranking module further refines candidate rankings by incorporating the intent compatibility between queries and legal articles. We evaluate LexPath on three Chinese benchmarks spanning diverse query scenarios and further assess its cross-jurisdictional applicability on a Japanese benchmark. Experimental results demonstrate that LexPath consistently outperforms lexical, dense, hybrid, and adaptive retrieval-augmented generation (RAG) baselines across all four benchmarks. Ablation studies further verify the effectiveness of each component. The data and code are available at https://github.com/wxliuabigail/LexPath.

cs.IR↗

arXiv2026

PhysEditBench: A Protocol-Conditioned Benchmark for Dense Physical-Map Prediction with Image Editors

Can general-purpose image editors predict physical maps from a single RGB image? General-purpose image editors differ from standard task-specific dense-prediction models: they do not directly take an image and output a physical map. Instead, they must be guided by prompts, examples, or image-based textual cues. To this end, we introduce PhysEditBench, a novel protocol-conditioned benchmark to evaluate and standardize image editors in dense physical-map prediction that covers five targets: depth, normal, albedo, roughness, and metallic maps. For evaluation data, we build a target-dependent benchmark substrate. We use OpenRooms-FF for depth, surface normal, albedo, and roughness, InteriorVerse as an additional source for depth, normal, albedo, and a new procedurally generated source for metallic maps. We curate the data with quality checks, valid-region masks, scene-level sampling, and lighting-based stress subsets to ensure reliable and diverse evaluation. For each target, PhysEditBench defines a fixed protocol that specifies the allowed input, expected output format, and scoring procedure. Each score, therefore, reflects the performance of a model under a specified protocol, rather than its best possible performance under all prompts or interaction modes. Experimental results show that specialized models remain much stronger on depth, normal, and albedo, and stronger image editors can produce more reasonable map-like outputs. For roughness and metallic, image editors can match or outperform specialized baselines on some scalar metrics, but they still suffer from structural errors, sparsity effects, and sensitivity to lighting.

cs.CV↗

arXiv2026

M3CoTBench: Benchmark Chain-of-Thought of MLLMs in Medical Image Understanding

Chain-of-Thought (CoT) reasoning has proven effective in enhancing large language models by encouraging step-by-step intermediate reasoning, and recent advances have extended this paradigm to Multimodal Large Language Models (MLLMs). In the medical domain, where diagnostic decisions depend on nuanced visual cues and sequential reasoning, CoT aligns naturally with clinical thinking processes. However, current benchmarks for medical image understanding generally focus on the final answer while ignoring the reasoning path. Such opaque reasoning processes lack reliable bases for judgment, making it difficult to assist doctors in diagnosis. To address this gap, we introduce a new M3CoTBench benchmark specifically designed to evaluate the correctness, efficiency, impact, and consistency of CoT reasoning in medical image understanding. M3CoTBench features 1) a diverse, multi-level difficulty dataset covering 24 examination types, 2) 13 varying-difficulty tasks, 3) a suite of CoT-specific evaluation metrics (correctness, efficiency, impact, and consistency) tailored to clinical reasoning, and 4) a performance analysis of multiple MLLMs. M3CoTBench systematically evaluates CoT reasoning across diverse medical imaging tasks, revealing current limitations of MLLMs in generating reliable and clinically interpretable reasoning, and aims to foster the development of transparent, trustworthy, and diagnostically accurate AI systems for healthcare. Project page at https://juntaojianggavin.github.io/projects/M3CoTBench/.

eess.IV↗

arXiv2025

UltraLBM-UNet: Ultralight Bidirectional Mamba-based Model for Skin Lesion Segmentation

Skin lesion segmentation is a crucial step in dermatology for guiding clinical decision-making. However, existing methods for accurate, robust, and resource-efficient lesion analysis have limitations, including low performance and high computational complexity. To address these limitations, we propose UltraLBM-UNet, a lightweight U-Net variant that integrates a bidirectional Mamba-based global modeling mechanism with multi-branch local feature perception. The proposed architecture integrates efficient local feature injection with bidirectional state-space modeling, enabling richer contextual interaction across spatial dimensions while maintaining computational compactness suitable for point-of-care deployment. Extensive experiments on the ISIC 2017, ISIC 2018, and PH2 datasets demonstrate that our model consistently achieves state-of-the-art segmentation accuracy, outperforming existing lightweight and Mamba counterparts with only 0.034M parameters and 0.060 GFLOPs. In addition, we introduce a hybrid knowledge distillation strategy to train an ultra-compact student model, where the distilled variant UltraLBM-UNet-T, with only 0.011M parameters and 0.019 GFLOPs, achieves competitive segmentation performance. These results highlight the suitability of UltraLBM-UNet for point-of-care deployment, where accurate and robust lesion analyses are essential. The source code is publicly available at https://github.com/LinLinLin-X/UltraLBM-UNet.

cs.CV↗

arXiv2025

RWKV-UNet: Improving UNet with Long-Range Cooperation for Effective Medical Image Segmentation

In recent years, significant advancements have been made in deep learning for medical image segmentation, particularly with convolutional neural networks (CNNs) and transformer models. However, CNNs face limitations in capturing long-range dependencies, while transformers suffer from high computational complexity. To address this, we propose RWKV-UNet, a novel model that integrates the RWKV (Receptance Weighted Key Value) structure into the U-Net architecture. This integration enhances the model's ability to capture long-range dependencies and to improve contextual understanding, which is crucial for accurate medical image segmentation. We build a strong encoder with developed Global-Local Spatial Perception (GLSP) blocks combining CNNs and RWKVs. We also propose a Cross-Channel Mix (CCM) module to improve skip connections with multi-scale feature fusion, achieving global channel information integration. Experiments on 11 benchmark datasets show that the RWKV-UNet achieves state-of-the-art performance on various types of medical image segmentation tasks. Additionally, smaller variants, RWKV-UNet-S and RWKV-UNet-T, balance accuracy and computational efficiency, making them suitable for broader clinical applications.

eess.IV↗

arXiv2025

BioNeuralNet: A Graph Neural Network based Multi-Omics Network Data Analysis Tool

Multi-omics data offer unprecedented insights into complex biological systems, yet their high dimensionality, sparsity, and intricate interactions pose significant analytical challenges. Network-based approaches have advanced multi-omics research by effectively capturing biologically relevant relationships among molecular entities. While these methods are powerful for representing molecular interactions, there remains a need for tools specifically designed to effectively utilize these network representations across diverse downstream analyses. To fulfill this need, we introduce BioNeuralNet, a flexible and modular Python framework tailored for end-to-end network-based multi-omics data analysis. BioNeuralNet leverages Graph Neural Networks (GNNs) to learn biologically meaningful low-dimensional representations from multi-omics networks, converting these complex molecular networks into versatile embeddings. BioNeuralNet supports all major stages of multi-omics network analysis, including several network construction techniques, generation of low-dimensional representations, and a broad range of downstream analytical tasks. Its extensive utilities, including diverse GNN architectures, and compatibility with established Python packages (e.g., scikit-learn, PyTorch, NetworkX), enhance usability and facilitate quick adoption. BioNeuralNet is an open-source, user-friendly, and extensively documented framework designed to support flexible and reproducible multi-omics network analysis in precision medicine.

cs.LG↗

arXiv2025

The Influence of Initial Connectivity on Biologically Plausible Learning

Understanding how the brain learns can be advanced by investigating biologically plausible learning rules -- those that obey known biological constraints, such as locality, to serve as valid brain learning models. Yet, many studies overlook the role of architecture and initial synaptic connectivity in such models. Building on insights from deep learning, where initialization profoundly affects learning dynamics, we ask a key but underexplored neuroscience question: how does initial synaptic connectivity shape learning in neural circuits? To investigate this, we train recurrent neural networks (RNNs), which are widely used for brain modeling, with biologically plausible learning rules. Our findings reveal that initial weight magnitude significantly influences the learning performance of such rules, mirroring effects previously observed in training with backpropagation through time (BPTT). By examining the maximum Lyapunov exponent before and after training, we uncovered the greater demands that certain initialization schemes place on training to achieve desired information propagation properties. Consequently, we extended the recently proposed gradient flossing method, which regularizes the Lyapunov exponents, to biologically plausible learning and observed an improvement in learning performance. To our knowledge, we are the first to examine the impact of initialization on biologically plausible learning rules for RNNs and to subsequently propose a biologically plausible remedy. Such an investigation can lead to neuroscientific predictions about the influence of initial connectivity on learning dynamics and performance, as well as guide neuromorphic design.

cs.NE↗