Lei Che
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Lei Che has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Chengcong Wu
College of Computer Science, Beijing Information Science and Technology University, Beijing 102206, China
1 shared publication Fan Jiang
University of Pittsburgh, America
1 shared publicationHongrui Zhang
Beijing Information Science and Technology University
1 shared publication- Junling Ren
Beijing Information Science and Technology University
1 shared publication Leilei Li
Beijing Information Science and Technology University
1 shared publicationXiao Ping Yang
1 shared publication
Latest Publications
- 30 August 2025
- Published by Springer Nature
- p. 187-195
Research on Night Cloud Amount Calculation Based on Transfer Learning
- 30 August 2025
- Published by Springer Nature
- p. 187-195
Research on Night Cloud Amount Calculation Based on Transfer Learning
- Open Access
- 31 May 2025
- Published by MDPI AG in Electronics
- Vol. 14 (11) , 2267
Large Language Model Text Adversarial Defense Method Based on Disturbance Detection and Error Correction
This study aims to effectively defend against complex and diverse adversarial attacks, enhance adversarial defense performance in the field of textual adversarial examples, and address existing issues in current defense methods such as models’ inability to accurately detect perturbations and perform effective error correction. To this end, we propose a Large Language Model Adversarial Defense (LLMAD) method based on perturbation detection and correction. The LLMAD framework consists of two modules: a perturbation detection module and a perturbation correction module. The detection module combines pre-trained models with soft-masking layers to rapidly determine whether samples contain adversarial perturbations. The correction module employs large language models fine-tuned using an adversarial example perturbation dataset constructed through data augmentation, enhancing task performance in adversarial defense and enabling efficient and accurate error correction of adversarial samples. Experimental results demonstrate that, across multiple datasets, various target models defended through the LLMAD method achieve satisfactory defense effectiveness against different types of adversarial attacks. The classification accuracy of defended models shows an average improvement of 66.8% compared to undefended baselines and outperforms existing methods by 13.4% on average. Additional perturbation detection performance tests and ablation studies further validate the accuracy of detection capabilities and the effectiveness of module combinations. A series of experiments confirm that the LLMAD method can significantly enhance defensive effectiveness against textual adversarial attacks.
- 31 May 2025
- Published by MDPI AG in Electronics
- Vol. 14 (11) , 2267
Open AccessLarge Language Model Text Adversarial Defense Method Based on Disturbance Detection and Error Correction
This study aims to effectively defend against complex and diverse adversarial attacks, enhance adversarial defense performance in the field of textual adversarial examples, and address existing issues in current defense methods such as models’ inability to accurately detect perturbations and perform effective error correction. To this end, we propose a Large Language Model Adversarial Defense (LLMAD) method based on perturbation detection and correction. The LLMAD framework consists of two modules: a perturbation detection module and a perturbation correction module. The detection module combines pre-trained models with soft-masking layers to rapidly determine whether samples contain adversarial perturbations. The correction module employs large language models fine-tuned using an adversarial example perturbation dataset constructed through data augmentation, enhancing task performance in adversarial defense and enabling efficient and accurate error correction of adversarial samples. Experimental results demonstrate that, across multiple datasets, various target models defended through the LLMAD method achieve satisfactory defense effectiveness against different types of adversarial attacks. The classification accuracy of defended models shows an average improvement of 66.8% compared to undefended baselines and outperforms existing methods by 13.4% on average. Additional perturbation detection performance tests and ablation studies further validate the accuracy of detection capabilities and the effectiveness of module combinations. A series of experiments confirm that the LLMAD method can significantly enhance defensive effectiveness against textual adversarial attacks.
- Open Access
- 8 March 2017
- Published by EDP Sciences in MATEC Web of Conferences
- Vol. 100, 05001
Application and research on business intelligence in audit business
MATEC Web of Conferences, open access proceedings in Materials science, Engineering and Chemistry
- 8 March 2017
- Published by EDP Sciences in MATEC Web of Conferences
- Vol. 100, 05001
Open AccessApplication and research on business intelligence in audit business
MATEC Web of Conferences, open access proceedings in Materials science, Engineering and Chemistry
- 1 February 2014
- Published by Trans Tech Publications, Ltd. in Applied Mechanics and Materials
- Vol. 519-520, 456-459
Research and Application of Mobile Agent in E-Commerce System
- 1 February 2014
- Published by Trans Tech Publications, Ltd. in Applied Mechanics and Materials
- Vol. 519-520, 456-459
Research and Application of Mobile Agent in E-Commerce System