Toward Fine-Grained Forgetting:Attribute Unlearning for Multimodal Large Language Models

arXiv:2608.01008 · cs.AI · Submitted 2026-08-02 · Read on arXiv

cs.AI

Submitted: 2026-08-02

Updated: 2026-10-12

Terminology

• Benchmark — 2× in this paper · explained in 2 episodes
• Activation Patching — 1× in this paper · explained in 2 episodes
• Causal Localization and Retain-Aware Projection — 1× in this paper
• Machine Unlearning — 1× in this paper · explained in 3 episodes
• Training-Free Framework — 1× in this paper · explained in 2 episodes

Sources

• One Modality to Forget Them All: Enhancing Cross-Modal Unlearning in Vision-Language Models
• Cross-Modal Attention Guided Unlearning in Vision-Language Models
• Machine Unlearning
• Targeted Unlearning with Single Layer Unlearning Gradient
• Cross-Modal Safety Alignment: Is textual unlearning all you need?
• Visual-Noise Guided In-Context Distillation for Multimodal Large Language Model Unlearning
• MultiDelete for Multimodal Machine Unlearning
• MLLMEraser: Achieving Test-Time Unlearning in Multimodal Large Language Models through Activation Steering
• SineProject: Machine Unlearning for Stable Vision Language Alignment
• PPU-Bench:Real World Benchmark for Personalized Partial Unlearning in Vision Language Models
• Object Hallucination-Free Reinforcement Unlearning for Vision-Language Models
• PULSE: Practical Evaluation Scenarios for Large Multimodal Model Unlearning
• Knowledge Vector Weakening: Efficient Training-free Unlearning for Large Vision-Language Models
• Erase Persona, Forget Lore: Benchmarking Multimodal Copyright Unlearning in Large Vision Language Models
• Cross-Modal Unlearning via Influential Neuron Path Editing in Multimodal Large Language Models
• Machine Unlearning: Taxonomy, Metrics, Applications, Challenges, and Prospects
• Continual Learning and Private Unlearning
• MIRROR: Manifold Ideal Reference ReconstructOR for Generalizable AI-Generated Image Detection
• Protecting Privacy in Multimodal Large Language Models with MLLMU-Bench
• TOFU: A Task of Fictitious Unlearning for LLMs

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
• Agents' Last Exam
• 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