search

VLM-Attack

Circumventing Concept Erasure Methods For Text-to-Image Generative Modelschevron-rightEfficient LLM-Jailbreaking by Introducing Visual Modalitychevron-rightFrom LLMs to MLLMs: Exploring the Landscape of Multimodal Jailbreakingchevron-rightAdversarial Attacks on Multimodal Agentschevron-rightVisual-RolePlay: Universal Jailbreak Attack on MultiModal Large Language Models via Role-playing Imachevron-rightCross-Modality Jailbreak and Mismatched Attacks on Medical Multimodal Large Language Modelschevron-rightTypography Leads Semantic Diversifying: Amplifying Adversarial Transferability across Multimodal Larchevron-rightWhite-box Multimodal Jailbreaks Against Large Vision-Language Modelschevron-rightRed Teaming Visual Language Modelschevron-rightPrivate Attribute Inference from Images with Vision-Language Modelschevron-rightAssessment of Multimodal Large Language Models in Alignment with Human Valueschevron-rightPrivacy-Aware Visual Language Modelschevron-rightLearning To See But Forgetting To Follow: Visual Instruction Tuning Makes LLMs More Prone To Jailbrechevron-rightVision-LLMs Can Fool Themselves with Self-Generated Typographic Attackschevron-rightRed Teaming Visual Language Modelschevron-rightAdversarial Illusions in Multi-Modal Embeddingschevron-rightUniversal Prompt Optimizer for Safe Text-to-Image Generationchevron-rightOn the Proactive Generation of Unsafe Images From Text-To-Image Models Using Benign Promptschevron-rightAdversarial Illusions in Multi-Modal Embeddingschevron-rightStop Reasoning! When Multimodal LLMs with Chain-of-Thought Reasoning Meets Adversarial Imageschevron-rightINSTRUCTTA: Instruction-Tuned Targeted Attack for Large Vision-Language Modelschevron-rightOn the Robustness of Large Multimodal Models Against Image Adversarial Attackschevron-rightHijacking Context in Large Multi-modal Modelschevron-rightTransferable Multimodal Attack on Vision-Language Pre-training Modelschevron-rightImages are Achilles’ Heel of Alignment: Exploiting Visual Vulnerabilities for Jailbreaking Multimodachevron-rightAN IMAGE IS WORTH 1000 LIES: ADVERSARIAL TRANSFERABILITY ACROSS PROMPTS ON VISIONLANGUAGE MODELSchevron-rightTest-Time Backdoor Attacks on Multimodal Large Language Modelschevron-rightJAILBREAK IN PIECES: COMPOSITIONAL ADVERSARIAL ATTACKS ON MULTI-MODAL LANGUAGE MODELSchevron-rightJailbreaking Attack against Multimodal Large Language Modelchevron-rightJailbreaking GPT-4V via Self-Adversarial Attacks with System Promptschevron-rightIMAGE HIJACKS: ADVERSARIAL IMAGES CAN CONTROL GENERATIVE MODELS AT RUNTIMEchevron-rightVISUAL ADVERSARIAL EXAMPLES JAILBREAK ALIGNED LARGE LANGUAGE MODELSchevron-rightVision-LLMs Can Fool Themselves with Self-Generated Typographic Attackschevron-rightQuery-Relevant Images Jailbreak Large Multi-Modal Modelschevron-rightTowards Adversarial Attack on Vision-Language Pre-training Modelschevron-rightHowMany Are Unicorns in This Image? ASafety Evaluation Benchmark for Vision LLMschevron-rightSA-Attack: Improving Adversarial Transferability of Vision-Language Pre-training Models via Self-Auchevron-rightMISUSING TOOLS IN LARGE LANGUAGE MODELS WITH VISUAL ADVERSARIAL EXAMPLESchevron-rightVL-Trojan: Multimodal Instruction Backdoor Attacks against Autoregressive Visual Language Modelschevron-rightINSTRUCTTA: Instruction-Tuned Targeted Attack for Large Vision-Language Modelschevron-rightSet-level Guidance Attack: Boosting Adversarial Transferability of Vision-Language Pre-training Modchevron-rightShadowcast: STEALTHY DATA POISONING ATTACKS AGAINST VISION-LANGUAGE MODELSchevron-rightFigStep: Jailbreaking Large Vision-language Models via Typographic Visual Promptschevron-rightTHE WOLF WITHIN: COVERT INJECTION OF MALICE INTO MLLM SOCIETIES VIA AN MLLM OPERATIVEchevron-rightStop Reasoning! When Multimodal LLMs with Chain-of-Thought Reasoning Meets Adversarial Imageschevron-rightAgent Smith: A Single Image Can Jailbreak One Million Multimodal LLM Agents Exponentially Fastchevron-rightHow Robust is Google’s Bard to Adversarial Image Attacks?chevron-rightOnEvaluating Adversarial Robustness of Large Vision-Language Modelschevron-rightOnthe Adversarial Robustness of Multi-Modal Foundation Modelschevron-rightAre aligned neural networks adversarially aligned?chevron-rightREADING ISN’T BELIEVING: ADVERSARIAL ATTACKS ON MULTI-MODAL NEURONSchevron-rightBlack Box Adversarial Prompting for Foundation Modelschevron-rightEvaluation and Analysis of Hallucination in Large Vision-Language Modelschevron-rightFOOL YOUR (VISION AND) LANGUAGE MODEL WITH EMBARRASSINGLY SIMPLE PERMUTATIONSchevron-rightVL-Trojan: Multimodal Instruction Backdoor Attacks against Autoregressive Visual Language Modelschevron-rightTransferable Multimodal Attack on Vision-Language Pre-training Modelschevron-rightBadCLIP: Dual-Embedding Guided Backdoor Attack on Multimodal Contrastive Learningchevron-rightAdvCLIP: Downstream-agnostic Adversarial Examples in Multimodal Contrastive Learningchevron-right
PreviousScalable Performance Analysis for Vision-Language ModelsNextCircumventing Concept Erasure Methods For Text-to-Image Generative Models

Last updated