Imagen de Portada

TokenSwap: Backdoor Attack on the Compositional Understanding of Large Vision-Language Models

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Zhang, Zhifang, Tao, Qiqi, Lv, Jiaqi, Zhao, Na, Feng, Lei, Zhou, Joey Tianyi
Formato:	Preprint
Publicado:	2025
Materias:	Computer Vision and Pattern Recognition
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Evolving from Single-modal to Multi-modal Facial Deepfake Detection: Progress and Challenges
por: Liu, Ping, et al.
Publicado: (2024)

Test-Time Attention Purification for Backdoored Large Vision Language Models
por: Zhang, Zhifang, et al.
Publicado: (2026)

Revisiting Backdoor Attacks against Large Vision-Language Models from Domain Shift
por: Liang, Siyuan, et al.
Publicado: (2024)

PointBA: Towards Backdoor Attacks in 3D Point Cloud
por: Li, Xinke, et al.
Publicado: (2021)

BackdoorVLM: A Benchmark for Backdoor Attacks on Vision-Language Models
por: Li, Juncheng, et al.
Publicado: (2025)

Mitigating Entangled Steering in Large Vision-Language Models for Hallucination Reduction
por: Zhang, Yuanhong, et al.
Publicado: (2026)

Backdoor Attack on Vision Language Models with Stealthy Semantic Manipulation
por: Zhong, Zhiyuan, et al.
Publicado: (2025)

MTAttack: Multi-Target Backdoor Attacks against Large Vision-Language Models
por: Wang, Zihan, et al.
Publicado: (2025)

Stealthy Backdoor Attack in Self-Supervised Learning Vision Encoders for Large Vision Language Models
por: Liu, Zhaoyi, et al.
Publicado: (2025)

Defending Multimodal Backdoored Models by Repulsive Visual Prompt Tuning
por: Zhang, Zhifang, et al.
Publicado: (2024)

Natural Reflection Backdoor Attack on Vision Language Model for Autonomous Driving
por: Liu, Ming, et al.
Publicado: (2025)

TrojVLM: Backdoor Attack Against Vision Language Models
por: Lyu, Weimin, et al.
Publicado: (2024)

Data-independent Module-aware Pruning for Hierarchical Vision Transformers
por: He, Yang, et al.
Publicado: (2024)

AttackVLA: Benchmarking Adversarial and Backdoor Attacks on Vision-Language-Action Models
por: Li, Jiayu, et al.
Publicado: (2025)

Modest-Align: Data-Efficient Alignment for Vision-Language Models
por: Liu, Jiaxiang, et al.
Publicado: (2025)

KPL: Training-Free Medical Knowledge Mining of Vision-Language Models
por: Liu, Jiaxiang, et al.
Publicado: (2025)

SCOPE: Saliency-Coverage Oriented Token Pruning for Efficient Multimodel LLMs
por: Deng, Jinhong, et al.
Publicado: (2025)

Learning to Detect Unseen Jailbreak Attacks in Large Vision-Language Models
por: Liang, Shuang, et al.
Publicado: (2025)

Hulu-Med: A Transparent Generalist Model towards Holistic Medical Vision-Language Understanding
por: Jiang, Songtao, et al.
Publicado: (2025)

Multi-Cue Adaptive Visual Token Pruning for Large Vision-Language Models
por: Luan, Bozhi, et al.
Publicado: (2025)

Ablate-to-Validate: Are Vision-Language Models Really Using Continuous Thought Tokens?
por: Zhang, Tianyi, et al.
Publicado: (2026)

Improving Generalizability and Undetectability for Targeted Adversarial Attacks on Multimodal Pre-trained Models
por: Zhang, Zhifang, et al.
Publicado: (2025)

Backdooring Vision-Language Models with Out-Of-Distribution Data
por: Lyu, Weimin, et al.
Publicado: (2024)

FlowHijack: A Dynamics-Aware Backdoor Attack on Flow-Matching Vision-Language-Action Models
por: An, Xinyuan, et al.
Publicado: (2026)

NullSwap: Proactive Identity Cloaking Against Deepfake Face Swapping
por: Wang, Tianyi, et al.
Publicado: (2025)

Act in Collusion: Distributed Multi-Target Backdoor Attacks in Federated Learning
por: Liu, Tao, et al.
Publicado: (2024)

Review of Hallucination Understanding in Large Language and Vision Models
por: Ho, Zhengyi, et al.
Publicado: (2025)

Guiding a Diffusion Model by Swapping Its Tokens
por: Zhang, Weijia, et al.
Publicado: (2026)

Multi-Turn Adaptive Prompting Attack on Large Vision-Language Models
por: Choi, In Chong, et al.
Publicado: (2026)

Physical Backdoor: Towards Temperature-based Backdoor Attacks in the Physical World
por: Yin, Wen, et al.
Publicado: (2024)

ZipVL: Efficient Large Vision-Language Models with Dynamic Token Sparsification
por: He, Yefei, et al.
Publicado: (2024)

Modelship Attribution: Tracing Multi-Stage Manipulations Across Generative Models
por: Tan, Zhiya, et al.
Publicado: (2025)

Goal-oriented Backdoor Attack against Vision-Language-Action Models via Physical Objects
por: Zhou, Zirun, et al.
Publicado: (2025)

LanP: Rethinking the Impact of Language Priors in Large Vision-Language Models
por: Wu, Zongyu, et al.
Publicado: (2025)

EM-KD: Distilling Efficient Multimodal Large Language Model with Unbalanced Vision Tokens
por: Feng, Ze, et al.
Publicado: (2025)

A Glimpse to Compress: Dynamic Visual Token Pruning for Large Vision-Language Models
por: Zeng, Quan-Sheng, et al.
Publicado: (2025)

HybridToken-VLM: Hybrid Token Compression for Vision-Language Models
por: Zhang, Jusheng, et al.
Publicado: (2025)

Learning to Detect Unknown Jailbreak Attacks in Large Vision-Language Models
por: Liang, Shuang, et al.
Publicado: (2025)

Fewer Tokens and Fewer Videos: Extending Video Understanding Abilities in Large Vision-Language Models
por: Chen, Shimin, et al.
Publicado: (2024)

Variation-aware Vision Token Dropping for Faster Large Vision-Language Models
por: Chen, Junjie, et al.
Publicado: (2025)