Tao Yin

Gang Bao, Liwei Xu, Tao Yin

In this paper, we consider the boundary integral equation (BIE) method for solving the exterior Neumann boundary value problems of elastic and thermoelastic waves in three dimensions based on the Fredholm integral equations of the first kind. The innovative contribution of this work lies in the proposal of the new regularized formulations for the hyper-singular boundary integral operators (BIO) associated with the time-harmonic elastic and thermoelastic wave equations. With the help of the new regularized formulations, we only need to compute the integrals with weak singularities at most in the corresponding variational forms of the boundary integral equations. The accuracy of the regularized formulations is demonstrated through numerical examples using the Galerkin boundary element method (BEM).

Gang Bao, Tao Yin, Fang Zeng

This paper concerns the reconstruction of multiple elastic parameters (Lamé parameters and density) of an inhomogeneous medium embedded in an infinite homogeneous isotropic background in $\mathbb{R}^2$. The direct scattering problem is reduced to an equivalent system on a bounded domain by introducing an exact transparent boundary condition and the wellposedness of the corresponding variational problem is established. The Fréchet differentiability of the near-field scattering map is studied with respect to the elastic parameters. Based on the multi-frequency measurement data and its phaseless term, two Landweber iterative algorithms are developed for the reconstruction of the multiple elastic parameters. Numerical examples, indicating that plane pressure incident wave is a better choice, are presented to show the validity and accuracy of our methods.

Oscar P. Bruno, Liwei Xu, Tao Yin

We present a novel approach for the numerical solution of problems of elastic scattering by open arcs in two dimensions. Our methodology relies on the composition of weighted versions of the classical operators associated with Dirichlet and Neumann boundary conditions in conjunction with a certain "open-arc elastic Calderón relation" whose validity is demonstrated in this paper on the basis of numerical experiments, but whose rigorous mathematical proof is left for future work. Using this Calderón relation in conjunction with spectrally accurate quadrature rules and the Krylov-subspace linear algebra solver GMRES, the proposed overall open-arc elastic solver produces results of high accuracy in small number of iterations---for low and high frequencies alike. A variety of numerical examples in this paper demonstrate the accuracy and efficiency of the proposed methodology.

Liwei Xu, Tao Yin

It is well known that the Fourier series Dirichlet-to-Neumann (DtN) boundary condition can be used to solve the Helmholtz equation in unbounded domains. In this work, applying such DtN boundary condition and using the finite element method, we solve and analyze a two dimensional transmission problem describing elastic waves inside a bounded and closed elastic obstacle and acoustic waves outside it. We are mainly interested in analyzing the DtN boundary condition of the Helmholtz equation in order to establish the well-posedness results of the approximated variational equation, and further derive a priori error estimates involving effects of both the finite element discretization and the DtN boundary condition truncation. Finally, some numerical results are presented to illustrate the accuracy of the numerical scheme.

Yuepeng Chen, Weiping Ding, Hengrong Ju et al.

Feature selection is a vital technique in machine learning, as it can reduce computational complexity, improve model performance, and mitigate the risk of overfitting. However, the increasing complexity and dimensionality of datasets pose significant challenges in the selection of features. Focusing on these challenges, this paper proposes a cascaded two-stage feature clustering and selection algorithm for fuzzy decision systems. In the first stage, we reduce the search space by clustering relevant features and addressing inter-feature redundancy. In the second stage, a clustering-based sequentially forward selection method that explores the global and local structure of data is presented. We propose a novel metric for assessing the significance of features, which considers both global separability and local consistency. Global separability measures the degree of intra-class cohesion and inter-class separation based on fuzzy membership, providing a comprehensive understanding of data separability. Meanwhile, local consistency leverages the fuzzy neighborhood rough set model to capture uncertainty and fuzziness in the data. The effectiveness of our proposed algorithm is evaluated through experiments conducted on 18 public datasets and a real-world schizophrenia dataset. The experiment results demonstrate our algorithm's superiority over benchmarking algorithms in both classification accuracy and the number of selected features.

Li Huang, Zhongxin Liu, Yifan Wu et al.

Large Language Models (LLMs) for code generation can replicate insecure patterns from their training data. To mitigate this, a common strategy for security hardening is to fine-tune models using supervision derived from the final transformer layer. However, this design may suffer from a final-layer bottleneck: vulnerability-discriminative cues can be distributed across layers and become less detectable near the output representations optimized for next-token prediction. To diagnose this issue, we perform layer-wise linear probing. We observe that vulnerability-related signals are most detectable in a band of intermediate-to-upper layers yet attenuate toward the final layers. Motivated by this observation, we introduce DeepGuard, a framework that leverages distributed security-relevant cues by aggregating representations from multiple upper layers via an attention-based module. The aggregated signal powers a dedicated security analyzer within a multi-objective training objective that balances security enhancement and functional correctness, and further supports a lightweight inference-time steering strategy. Extensive experiments across five code LLMs demonstrate that DeepGuard improves the secure-and-correct generation rate by an average of 11.9% over strong baselines such as SVEN. It also preserves functional correctness while exhibiting generalization to held-out vulnerability types. Our code is public at https://github.com/unknownhl/DeepGuard.

Shen Li, Li Huang, Shaoxiong Zhan et al.

Large language models (LLMs) exhibit strong generative capabilities and have shown great potential in code generation. Existing chain-of-thought (CoT) prompting methods enhance model reasoning by eliciting intermediate steps, but suffer from two major limitations: First, their uniform application tends to induce overthinking on simple tasks. Second, they lack intention abstraction in code generation, such as explicitly modeling core algorithmic design and efficiency, leading models to focus on surface-level structures while neglecting the global problem objective. Inspired by the cognitive economy principle of engaging structured reasoning only when necessary to conserve cognitive resources, we propose RoutingGen, a novel difficulty-aware routing framework that dynamically adapts prompting strategies for code generation. For simple tasks, it adopts few-shot prompting; for more complex ones, it invokes a structured reasoning strategy, termed Intention Chain-of-Thought (ICoT), which we introduce to guide the model in capturing task intention, such as the core algorithmic logic and its time complexity. Experiments across three models and six standard code generation benchmarks show that RoutingGen achieves state-of-the-art performance in most settings, while reducing total token usage by 46.37% on average across settings. Furthermore, ICoT outperforms six existing prompting baselines on challenging benchmarks.

Tao Yin, Xiaohong Zhang, Jiacheng Zhang et al.

Effectively integrating molecular graph structures with Large Language Models (LLMs) is a key challenge in drug discovery. Most existing multi-modal alignment methods typically process these structures by fine-tuning the LLM or adding a static adapter simultaneously. However, these approaches have two main limitations: (1) it optimizes a shared parameter space across all molecular inputs, limiting the model's ability to capture instance-specific structural features; and (2) fine-tuning the LLM for molecular tasks can lead to catastrophic forgetting, undermining its general reasoning capabilities. In this paper, instead of static task-oriented adaptation, we propose an instance-specific parameter space alignment approach for each molecule on-the-fly. To this end, we introduce Molecule-aware Low-Rank Adaptation (MoRA) that produces a unique set of low-rank adaptation weights for each input molecular graph. These weights are then dynamically injected into a frozen LLM, allowing the model to adapt its reasoning to the structure of each molecular input, while preserving the LLM's core knowledge. Extensive experiments demonstrate that on key molecular tasks, such as chemical reaction prediction and molecular captioning, MoRA's instance-specific dynamic adaptation outperforms statically adapted baselines, including a 14.1% relative improvement in reaction prediction exact match and a 22% reduction in error for quantum property prediction. The code is available at https://github.com/jk-sounds/MoRA.

Tao Yin, Xiaohong Zhang, Shaochen Fu et al.

One main challenge in time series anomaly detection for industrial IoT lies in the complex spatio-temporal couplings within multivariate data. However, traditional anomaly detection methods focus on modeling spatial or temporal dependencies independently, resulting in suboptimal representation learning and limited sensitivity to anomalous dispersion in high-dimensional spaces. In this work, we conduct an empirical analysis showing that both normal and anomalous samples tend to scatter in high-dimensional space, especially anomalous samples are markedly more dispersed. We formalize this dispersion phenomenon as scattering, quantified by the mean pairwise distance among sample representations, and leverage it as an inductive signal to enhance spatio-temporal anomaly detection. Technically, we propose ScatterAD to model representation scattering across temporal and topological dimensions. ScatterAD incorporates a topological encoder for capturing graph-structured scattering and a temporal encoder for constraining over-scattering through mean squared error minimization between neighboring time steps. We introduce a contrastive fusion mechanism to ensure the complementarity of the learned temporal and topological representations. Additionally, we theoretically show that maximizing the conditional mutual information between temporal and topological views improves cross-view consistency and enhances more discriminative representations. Extensive experiments on multiple public benchmarks show that ScatterAD achieves state-of-the-art performance on multivariate time series anomaly detection. Code is available at this repository: https://github.com/jk-sounds/ScatterAD.

Fengxiao Tang, Huan Li, Ming Zhao et al.

Verifying the credibility of Cyber Threat Intelligence (CTI) is essential for reliable cybersecurity defense. However, traditional approaches typically treat this task as a static classification problem, relying on handcrafted features or isolated deep learning models. These methods often lack the robustness needed to handle incomplete, heterogeneous, or noisy intelligence, and they provide limited transparency in decision-making-factors that reduce their effectiveness in real-world threat environments. To address these limitations, we propose LRCTI, a Large Language Model (LLM)-based framework designed for multi-step CTI credibility verification. The framework first employs a text summarization module to distill complex intelligence reports into concise and actionable threat claims. It then uses an adaptive multi-step evidence retrieval mechanism that iteratively identifies and refines supporting information from a CTI-specific corpus, guided by LLM feedback. Finally, a prompt-based Natural Language Inference (NLI) module is applied to evaluate the credibility of each claim while generating interpretable justifications for the classification outcome. Experiments conducted on two benchmark datasets, CTI-200 and PolitiFact show that LRCTI improves F1-Macro and F1-Micro scores by over 5%, reaching 90.9% and 93.6%, respectively, compared to state-of-the-art baselines. These results demonstrate that LRCTI effectively addresses the core limitations of prior methods, offering a scalable, accurate, and explainable solution for automated CTI credibility verification

Tao Yin, Chen Zhao, Xiaoyan Liu et al.

Graph neural networks (GNNs) are proven effective in extracting complex node and structural information from graph data. While current GNNs perform well in node classification tasks within in-distribution (ID) settings, real-world scenarios often present distribution shifts, leading to the presence of out-of-distribution (OOD) nodes. OOD detection in graphs is a crucial and challenging task. Most existing research focuses on homogeneous graphs, but real-world graphs are often heterogeneous, consisting of diverse node and edge types. This heterogeneity adds complexity and enriches the informational content. To the best of our knowledge, OOD detection in heterogeneous graphs remains an underexplored area. In this context, we propose a novel methodology for OOD detection in heterogeneous graphs (OODHG) that aims to achieve two main objectives: 1) detecting OOD nodes and 2) classifying all ID nodes based on the first task's results. Specifically, we learn representations for each node in the heterogeneous graph, calculate energy values to determine whether nodes are OOD, and then classify ID nodes. To leverage the structural information of heterogeneous graphs, we introduce a meta-path-based energy propagation mechanism and an energy constraint to enhance the distinction between ID and OOD nodes. Extensive experimental findings substantiate the simplicity and effectiveness of OODHG, demonstrating its superiority over baseline models in OOD detection tasks and its accuracy in ID node classification.

Gang Bao, Liwei Xu, Tao Yin

This paper is concerned with a Galerkin boundary element method solving the two dimensional exterior elastic wave scattering problem. The original problem is first reduced to the so-called Burton-Miller (\cite{BM71}) boundary integral formulation, and essential mathematical features of its variational form are discussed. In numerical implementations, a newly-derived and analytically accurate regularization formula (\cite{YHX}) is employed for the numerical evaluation of hyper-singular boundary integral operator. A new computational approach is employed based on the series expansions of Hankel functions for the computation of weakly-singular boundary integral operators during the reduction of corresponding Galerkin equations into a discrete linear system. The effectiveness of proposed numerical methods is demonstrated using several numerical examples.

Gang Bao, Guanghui Hu, Yavar Kian et al.

We are concerned with time-dependent inverse source problems in elastodynamics. The source term is supposed to be the product of a spatial function and a temporal function with compact support. We present frequency-domain and time-domain approaches to show uniqueness in determining the spatial function from wave fields on a large sphere over a finite interval. Stability estimate of the temporal function from the data of one receiver and uniqueness result using partial boundary data are proved. Our arguments rely heavily on the use of the Fourier transform, which motivated inversion schemes that can be easily implemented. A Landweber iterative algorithm for recovering the spatial function and a non-iterative inversion scheme based on the uniqueness proof for recovering the temporal function are proposed. Numerical examples are demonstrated in both two and three dimensions.

Tao Yin, George C. Hsiao, Liwei Xu

This paper is concerned with boundary integral equation methods for solving the two-dimensional fluid-solid interaction problem. We reduce the problem to three differential systems of boundary integral equations via direct and indirect approaches. Existence and uniqueness results for variational solutions of boundary integral equations equations are established. Since in all these boundary variational formulations, the hypersingular boundary integral operator associated with the time-harmonic Navier equation is a dominated integral operator, we also include a new regularization formulation for this hypersingular operator, which allows us to treat the hypersingular kernel by a wealkly singular kernel. Numerical examples are presented to verify and validate the theoretical results.