الباحثون

Zhifang Zhang

المنشورات 2

نسخة أولية وصول مفتوح

Purifying Backdoored Large Vision-Language Models by Removing Hijacked Directions

Bojun Yang, Haochen Zhou, Zhifang Zhang وآخرون · 2026

Large vision-language models (LVLMs) are increasingly deployed in safety-critical applications, yet they remain vulnerable to backdoor attacks. Defending against such attacks remains costly, as existing methods require either extensive retraining on clean data or per-query intervention at inference time. To address thi …

نسخة أولية وصول مفتوح

Selective Channel Restoration for Backdoored Vision-Language Models

Shuming Liu, Zhifang Zhang, Suqin Yuan وآخرون · 2026

Vision-language models (VLMs) exhibit strong multimodal capabilities but remain vulnerable to backdoors implanted through poisoned fine-tuning data. Existing defenses often require extensive parameter updates during fine-tuning or incur per-query overhead during inference. To address these limitations, we propose Pertu …

المؤلفون المشاركون