Image de couverture de livre

When World Models Dream Wrong: Physical-Conditioned Adversarial Attacks against World Models

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Guo, Zhixiang, Liang, Siyuan, Balogh, Andras, Lunberry, Noah, Tu, Rong-Cheng, Jelasity, Mark, Tao, Dacheng
Format:	Preprint
Publié:	2026
Sujets:	Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Verification of the Implicit World Model in a Generative Model via Adversarial Sequences
par: Balogh, András, et autres
Publié: (2026)

How not to Stitch Representations to Measure Similarity: Task Loss Matching versus Direct Matching
par: Balogh, András, et autres
Publié: (2024)

CopyrightShield: Enhancing Diffusion Model Security against Copyright Infringement Attacks
par: Guo, Zhixiang, et autres
Publié: (2024)

CtrlAttack: A Unified Attack on World-Model Control in Diffusion Models
par: Xu, Shuhan, et autres
Publié: (2026)

No Soundness in the Real World: On the Challenges of the Verification of Deployed Neural Networks
par: Szász, Attila, et autres
Publié: (2025)

Self-supervised Adversarial Training of Monocular Depth Estimation against Physical-World Attacks
par: Cheng, Zhiyuan, et autres
Publié: (2024)

Black-Box Adversarial Attack on Vision Language Models for Autonomous Driving
par: Wang, Lu, et autres
Publié: (2025)

Revisiting Backdoor Attacks against Large Vision-Language Models from Domain Shift
par: Liang, Siyuan, et autres
Publié: (2024)

Evaluating the Adversarial Robustness of Semantic Segmentation: Trying Harder Pays Off
par: Halmosi, Levente, et autres
Publié: (2024)

ATAC: Augmentation-Based Test-Time Adversarial Correction for CLIP
par: Su, Linxiang, et autres
Publié: (2025)

Cert-LAS: Toward Certified Model Ownership Verification for Text-to-Image Diffusion Models via Layer-Adaptive Smoothing
par: Qi, Leyi, et autres
Publié: (2026)

Towards Physical World Backdoor Attacks against Skeleton Action Recognition
par: Zheng, Qichen, et autres
Publié: (2024)

Review of Hallucination Understanding in Large Language and Vision Models
par: Ho, Zhengyi, et autres
Publié: (2025)

Diffusion Model-Based Video Editing: A Survey
par: Sun, Wenhao, et autres
Publié: (2024)

Probabilistic Dreaming for World Models
par: Wong, Gavin
Publié: (2026)

ICLShield: Exploring and Mitigating In-Context Learning Backdoor Attacks
par: Ren, Zhiyao, et autres
Publié: (2025)

Who Generated This 3D Asset? Learning Source Attribution for Generative 3D Models
par: Ma, Sihan, et autres
Publié: (2026)

DreamWorld: Unified World Modeling in Video Generation
par: Tan, Boming, et autres
Publié: (2026)

ME: Trigger Element Combination Backdoor Attack on Copyright Infringement
par: Yang, Feiyu, et autres
Publié: (2025)

Natural Reflection Backdoor Attack on Vision Language Model for Autonomous Driving
par: Liu, Ming, et autres
Publié: (2025)

TRAP: Tail-aware Ranking Attack for World-Model Planning
par: Duan, Siyuan, et autres
Publié: (2026)

CogMorph: Cognitive Morphing Attacks for Text-to-Image Models
par: Jing, Zonglei, et autres
Publié: (2025)

Jailbreak Vision Language Models via Bi-Modal Adversarial Prompt
par: Ying, Zonghao, et autres
Publié: (2024)

A New World Framework: The Three Realms and Six Layers Model
par: Zhou, Dacheng
Publié: (2024)

Dream to Drive with Predictive Individual World Model
par: Gao, Yinfeng, et autres
Publié: (2025)

RoboDream: Compositional World Models for Scalable Robot Data Synthesis
par: Ye, Junjie, et autres
Publié: (2026)

Multi-granular Adversarial Attacks against Black-box Neural Ranking Models
par: Liu, Yu-An, et autres
Publié: (2024)

Visual Adversarial Attacks and Defenses in the Physical World: A Survey
par: Wei, Xingxing, et autres
Publié: (2022)

HarmonyDream: Task Harmonization Inside World Models
par: Ma, Haoyu, et autres
Publié: (2023)

Random Geometric Graphs in Reflexive Banach Spaces
par: Balogh, József, et autres
Publié: (2024)

UV-Attack: Physical-World Adversarial Attacks for Person Detection via Dynamic-NeRF-based UV Mapping
par: Li, Yanjie, et autres
Publié: (2025)

DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledge
par: Zhang, Wenyao, et autres
Publié: (2025)

ELBA-Bench: An Efficient Learning Backdoor Attacks Benchmark for Large Language Models
par: Liu, Xuxu, et autres
Publié: (2025)

Dreaming of Many Worlds: Learning Contextual World Models Aids Zero-Shot Generalization
par: Prasanna, Sai, et autres
Publié: (2024)

Do Neurons Dream of Primitive Operators? Wake-Sleep Compression Rediscovers Schank's Event Semantics
par: Balogh, Peter
Publié: (2026)

T2VShield: Model-Agnostic Jailbreak Defense for Text-to-Video Models
par: Liang, Siyuan, et autres
Publié: (2025)

Four Principles for Physically Interpretable World Models
par: Peper, Jordan, et autres
Publié: (2025)

SPAgent: Adaptive Task Decomposition and Model Selection for General Video Generation and Editing
par: Tu, Rong-Cheng, et autres
Publié: (2024)

Towards Transferable Targeted 3D Adversarial Attack in the Physical World
par: Huang, Yao, et autres
Publié: (2023)

CURLing the Dream: Contrastive Representations for World Modeling in Reinforcement Learning
par: Kich, Victor Augusto, et autres
Publié: (2024)