A new tractable Archimedean copula for full-range tail dependence
本文提出了一种新的阿基米德copula（FRA1），它具有闭式密度函数，能够捕捉上下尾部的全范围尾部相依性，并通过最大似然估计参数。
0 citationsRead paper
Selected work
Representative Papers
From Causal Plausibility to Causal Reliability: Evaluating LLMs as Calibrated Direct Causal-Edge Classifiers
研究评估了12种语言模型在因果关系判断上的可靠性，发现这些模型倾向于过度预测因果边，并且对直接因果关系的识别不够准确。传统的置信度估计不可靠，而跨提示和跨模型的一致性则提供了更好的信号。
0 citationsRead paper
Beyond Global Editing: Per-Instance Disentangled Subspaces for Training-Free Hallucination Mitigation in LVLMs
This work addresses the prevalent issue of hallucination in large vision-language models (LVLMs), where generated text often contradicts visual input, undermining model reliability. The authors propose a training-free, test-time hallucination suppression framework that dynamically constructs a disentangled hallucination subspace for each input instance and selectively suppresses the most likely hallucinatory directions via adaptive weighted projection, while preserving image-relevant semantics. Moving beyond conventional global editing strategies, this approach achieves instance-level dynamic modeling of hallucinations for the first time, substantially improving both suppression accuracy and generalization. Extensive experiments demonstrate consistent performance gains across diverse LVLM architectures and vision-language benchmarks, confirming the framework’s robustness, versatility, and efficiency.
0 citationsRead paper
A Physics-Informed Neural Network Framework for Elastodynamic Wave Propagation in Bimaterial Systems
This work investigates the reflection and transmission of transient elastic waves at the interface of a bimaterial system. A physics-informed neural network (PINN)-based surrogate model is proposed, which embeds the axisymmetric linear elastodynamic equations, initial and boundary conditions, and interfacial constraints directly into the loss function, and is trained using data from ANSYS explicit dynamics finite element simulations. To the best of our knowledge, this is the first application of PINNs to modeling elastic wave propagation in bimaterial systems under high strain rates, enabling generalization across time instances and material parameters without additional simulations. The model accurately reproduces displacement time histories, surface-averaged responses, and stress–strain evolution, showing excellent agreement with finite element results, and its robustness is confirmed through mesh sensitivity analysis.
0 citationsRead paper
From Theory to Practice: Code Generation Using LLMs for CAPEC and CWE Frameworks
This work addresses the critical gap in existing vulnerability datasets—the lack of multilingual code examples explicitly linked to CAPEC/CWE standards—which hinders both vulnerability comprehension and the development of robust security models. To bridge this gap, the study presents the first systematic integration of the CAPEC/CWE knowledge framework with large language model (LLM) generation techniques, leveraging GPT-4o, Llama, and Claude. Through carefully engineered prompts and a rigorous consistency validation mechanism, the authors construct a large-scale vulnerable code dataset encompassing 615 attack patterns across three major programming languages. The generated code exhibits high fidelity, with inter-model cosine similarity reaching 0.98 and strong accuracy, thereby providing a reliable foundation for training and evaluating vulnerability detection and repair models.
0 citationsRead paper
Recent publications
Latest Papers
A new tractable Archimedean copula for full-range tail dependence
本文提出了一种新的阿基米德copula（FRA1），它具有闭式密度函数，能够捕捉上下尾部的全范围尾部相依性，并通过最大似然估计参数。
0 citationsRead paper
From Causal Plausibility to Causal Reliability: Evaluating LLMs as Calibrated Direct Causal-Edge Classifiers
研究评估了12种语言模型在因果关系判断上的可靠性，发现这些模型倾向于过度预测因果边，并且对直接因果关系的识别不够准确。传统的置信度估计不可靠，而跨提示和跨模型的一致性则提供了更好的信号。
0 citationsRead paper
Beyond Global Editing: Per-Instance Disentangled Subspaces for Training-Free Hallucination Mitigation in LVLMs
This work addresses the prevalent issue of hallucination in large vision-language models (LVLMs), where generated text often contradicts visual input, undermining model reliability. The authors propose a training-free, test-time hallucination suppression framework that dynamically constructs a disentangled hallucination subspace for each input instance and selectively suppresses the most likely hallucinatory directions via adaptive weighted projection, while preserving image-relevant semantics. Moving beyond conventional global editing strategies, this approach achieves instance-level dynamic modeling of hallucinations for the first time, substantially improving both suppression accuracy and generalization. Extensive experiments demonstrate consistent performance gains across diverse LVLM architectures and vision-language benchmarks, confirming the framework’s robustness, versatility, and efficiency.
0 citationsRead paper
A Physics-Informed Neural Network Framework for Elastodynamic Wave Propagation in Bimaterial Systems
This work investigates the reflection and transmission of transient elastic waves at the interface of a bimaterial system. A physics-informed neural network (PINN)-based surrogate model is proposed, which embeds the axisymmetric linear elastodynamic equations, initial and boundary conditions, and interfacial constraints directly into the loss function, and is trained using data from ANSYS explicit dynamics finite element simulations. To the best of our knowledge, this is the first application of PINNs to modeling elastic wave propagation in bimaterial systems under high strain rates, enabling generalization across time instances and material parameters without additional simulations. The model accurately reproduces displacement time histories, surface-averaged responses, and stress–strain evolution, showing excellent agreement with finite element results, and its robustness is confirmed through mesh sensitivity analysis.
0 citationsRead paper
From Theory to Practice: Code Generation Using LLMs for CAPEC and CWE Frameworks
This work addresses the critical gap in existing vulnerability datasets—the lack of multilingual code examples explicitly linked to CAPEC/CWE standards—which hinders both vulnerability comprehension and the development of robust security models. To bridge this gap, the study presents the first systematic integration of the CAPEC/CWE knowledge framework with large language model (LLM) generation techniques, leveraging GPT-4o, Llama, and Claude. Through carefully engineered prompts and a rigorous consistency validation mechanism, the authors construct a large-scale vulnerable code dataset encompassing 615 attack patterns across three major programming languages. The generated code exhibits high fidelity, with inter-model cosine similarity reaching 0.98 and strong accuracy, thereby providing a reliable foundation for training and evaluating vulnerability detection and repair models.
0 citationsRead paper