Group Preference Collapse in Personalized Multimodal Large Language Models

arXiv:2607.22603 · cs.AI · Submitted 2026-06-12 · Read on arXiv

cs.AI

Submitted: 2026-06-12

Updated: 2026-09-30

Project page: https://prefmoe.github.io/ABSTRACT

Terminology

• Adaptation — 1× in this paper · explained in 1 episode
• LoRA Adaptation — 1× in this paper · explained in 2 episodes
• Personalization — 1× in this paper · explained in 2 episodes

Sources

• MC-LLaVA: Multi-Concept Personalized Vision-Language Model
• Online-PVLM: Advancing Personalized VLMs with Online Concept Learning
• Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
• M2A: Multimodal Memory Agent with Dual-Layer Hybrid Memory for Long-Term Personalized Interactions
• Jarvis: Towards Personalized AI Assistant via Personal KV-Cache Retrieval
• Small-Large Collaboration: Training-efficient Concept Personalization for Large VLM using a Meta Personalized Small VLM
• PEARL: Personalized Streaming Video Understanding Model
• Fine-Tuning Language Models from Human Preferences

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