CUSE: Fine-Grained Multimodal Unlearning with Counterfactual Guidance and On-Policy Self-Distillation
cs.AI, cs.CL
Submitted: 2026-10-10
Updated: 2026-10-10
Terminology
Sources
- Qwen2.5-VL Technical Report
- Cross-Modal Safety Alignment: Is textual unlearning all you need?
- MLLMEraser: Achieving Test-Time Unlearning in Multimodal Large Language Models through Activation Steering
- Who's Harry Potter? Approximate Unlearning in LLMs
- Reducing the Safety Tax in LLM Safety Alignment with On-Policy Self-Distillation
- ASRU: Activation Steering Meets Reinforcement Unlearning for Multimodal Large Language Models
- Direct Language Model Alignment from Online AI Feedback
- Distilling the Knowledge in a Neural Network
- Rethinking On-Policy Self-Distillation for Thinking Models
- MLUBench: A Benchmark for Lifelong Unlearning Evaluation in MLLMs
- Towards Reasoning-Preserving Unlearning in Multimodal Large Language Models
- TOFU: A Task of Fictitious Unlearning for LLMs
- GATES: Self-Distillation under Privileged Context with Consensus Gating
- Qwen3-VL Technical Report
- Guardrail Baselines for Unlearning in LLMs
- Black-Box On-Policy Distillation of Large Language Models
- Online Experiential Learning for Language Models
- On-Policy Context Distillation for Language Models
- Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning
- Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models
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