Cover Image

Visual CoT Makes VLMs Smarter but More Fragile

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Xu, Chunxue, Wang, Yiwei, Cai, Yujun, Hooi, Bryan, Li, Songze
Format:	Preprint
Published:	2025
Subjects:	Machine Learning Cryptography and Security
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

How does Watermarking Affect Visual Language Models in Document Understanding?
by: Xu, Chunxue, et al.
Published: (2025)

Tricking Retrievers with Influential Tokens: An Efficient Black-Box Corpus Poisoning Attack
by: Wang, Cheng, et al.
Published: (2025)

Enhancing Membership Inference Attacks on Diffusion Models from a Frequency-Domain Perspective
by: Lian, Puwei, et al.
Published: (2025)

Noise as a Probe: Membership Inference Attacks on Diffusion Models Leveraging Initial Noise
by: Lian, Puwei, et al.
Published: (2026)

FuncPoison: Poisoning Function Library to Hijack Multi-agent Autonomous Driving Systems
by: Long, Yuzhen, et al.
Published: (2025)

DeDe: Detecting Backdoor Samples for SSL Encoders via Decoders
by: Hou, Sizai, et al.
Published: (2024)

TUNI: A Textual Unimodal Detector for Identity Inference in CLIP Models
by: Li, Songze, et al.
Published: (2024)

OmniLytics+: A Secure, Efficient, and Affordable Blockchain Data Market for Machine Learning through Off-Chain Processing
by: Li, Songze, et al.
Published: (2024)

Noticing the Watcher: LLM Agents Can Infer CoT Monitoring from Blocking Feedback
by: Jiralerspong, Thomas, et al.
Published: (2026)

Graph Neural Network Explanations are Fragile
by: Li, Jiate, et al.
Published: (2024)

Verification of Machine Unlearning is Fragile
by: Zhang, Binchi, et al.
Published: (2024)

TrojanDam: Detection-Free Backdoor Defense in Federated Learning through Proactive Model Robustification utilizing OOD Data
by: Dai, Yanbo, et al.
Published: (2025)

Constructing Adversarial Examples for Vertical Federated Learning: Optimal Client Corruption through Multi-Armed Bandit
by: Yao, Duanyi, et al.
Published: (2024)

Awakening the Hydra: Stabilizing Multi-Concept Backdoor Injection in Text-to-Image Diffusion Models
by: Wang, Kai, et al.
Published: (2026)

Knowledge-Driven Multi-Turn Jailbreaking on Large Language Models
by: Li, Songze, et al.
Published: (2026)

Making Every Step Effective: Jailbreaking Large Vision-Language Models Through Hierarchical KV Equalization
by: Hao, Shuyang, et al.
Published: (2025)

Attack Smarter: Attention-Driven Fine-Grained Webpage Fingerprinting Attacks
by: Yuan, Yali, et al.
Published: (2025)

TooBadRL: Trigger Optimization to Boost Effectiveness of Backdoor Attacks on Deep Reinforcement Learning
by: Zhang, Mingxuan, et al.
Published: (2025)

URVFL: Undetectable Data Reconstruction Attack on Vertical Federated Learning
by: Yao, Duanyi, et al.
Published: (2024)

Attributing and Exploiting Safety Vectors through Global Optimization in Large Language Models
by: Chu, Fengheng, et al.
Published: (2026)

VLMGuard: Defending VLMs against Malicious Prompts via Unlabeled Data
by: Du, Xuefeng, et al.
Published: (2024)

Tit-for-Tat: Safeguarding Large Vision-Language Models Against Jailbreak Attacks via Adversarial Defense
by: Hao, Shuyang, et al.
Published: (2025)

On The Fragility of Benchmark Contamination Detection in Reasoning Models
by: Wang, Han, et al.
Published: (2025)

Turning Bias into Bugs: Bandit-Guided Style Manipulation Attacks on LLM Judges
by: Yang, Xianglin, et al.
Published: (2026)

Beauty and the Beast: Imperceptible Perturbations Against Diffusion-Based Face Swapping via Directional Attribute Editing
by: Huang, Yilong, et al.
Published: (2026)

Federated Learning for Cross-Domain Data Privacy: A Distributed Approach to Secure Collaboration
by: Zhang, Yiwei, et al.
Published: (2025)

Continuous Multi-Task Pre-training for Malicious URL Detection and Webpage Classification
by: Li, Yujie, et al.
Published: (2024)

Odysseus: Jailbreaking Commercial Multimodal LLM-integrated Systems via Dual Steganography
by: Li, Songze, et al.
Published: (2025)

Hidden Ads: Behavior Triggered Semantic Backdoors for Advertisement Injection in Vision Language Models
by: Yao, Duanyi, et al.
Published: (2026)

Automated Phishing Detection Using URLs and Webpages
by: Wang, Huilin, et al.
Published: (2024)

MIRAGE: Multimodal Immersive Reasoning and Guided Exploration for Red-Team Jailbreak Attacks
by: You, Wenhao, et al.
Published: (2025)

Is the Trigger Essential? A Feature-Based Triggerless Backdoor Attack in Vertical Federated Learning
by: Liu, Yige, et al.
Published: (2026)

Indiscriminate Data Poisoning Attacks on Neural Networks
by: Lu, Yiwei, et al.
Published: (2022)

Reasoning Introduces New Poisoning Attacks Yet Makes Them More Complicated
by: Foerster, Hanna, et al.
Published: (2025)

DMS: Addressing Information Loss with More Steps for Pragmatic Adversarial Attacks
by: Zhu, Zhiyu, et al.
Published: (2024)

Stealthy Yet Effective: Distribution-Preserving Backdoor Attacks on Graph Classification
by: Wang, Xiaobao, et al.
Published: (2025)

Whisper Smarter, not Harder: Adversarial Attack on Partial Suppression
by: Wong, Zheng Jie, et al.
Published: (2025)

No More, No Less: Least-Privilege Language Models
by: Rauba, Paulius, et al.
Published: (2026)

Toward More Generalized Malicious URL Detection Models
by: Tsai, YunDa, et al.
Published: (2022)

Deferred Poisoning: Making the Model More Vulnerable via Hessian Singularization
by: He, Yuhao, et al.
Published: (2024)