The Plot Twist: Jailbreaking Unified Multimodal Models with a Three-Act NarrativeAttack
cs.AI
Submitted: 2025-09-30
Updated: 2026-09-25
Terminology
Sources
- Emerging Properties in Unified Multimodal Pretraining
- Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling
- Rethinking Bottlenecks in Safety Fine-Tuning of Vision Language Models
- MetaMorph: Multimodal Understanding and Generation via Instruction Tuning
- AnyGPT: Unified Multimodal LLM with Discrete Sequence Modeling
- FlipAttack: Jailbreak LLMs via Flipping
- A Wolf in Sheep's Clothing: Generalized Nested Jailbreak Prompts can Fool Large Language Models Easily
- X-Teaming: Multi-Turn Jailbreaks and Defenses with Adaptive Multi-Agents
- AutoDAN-Turbo: A Lifelong Agent for Strategy Self-Exploration to Jailbreak LLMs
- Rewrite to Jailbreak: Discover Learnable and Transferable Implicit Harmfulness Instruction
- HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
- Universal and Transferable Adversarial Attacks on Aligned Language Models
- Seed-X: Building Strong Multilingual Translation LLM with 7B Parameters
- LMFusion: Adapting Pretrained Language Models for Multimodal Generation
- ILLUME+: Illuminating Unified MLLM with Dual Visual Tokenization and Diffusion Refinement
- EasyJailbreak: A Unified Framework for Jailbreaking Large Language Models
- TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation
- MUSE-VL: Modeling Unified VLM through Semantic Discrete Encoding
- ReasonGen-R1: CoT for Autoregressive Image generation models through SFT and RL
- Adaptive Group Policy Optimization: Towards Stable Training and Token-Efficient Reasoning
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
- VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection