SOURCE-LINKED INTELLIGENCE

A Visual Dependence-Aware Framework for Multimodal Unsupervised Continual Post-Training

arXiv · AI, language, vision and robotics · article · Aug 26, 2026 · UTC

In this paper, we explore a novel task of Multimodal Unsupervised Continual Post-Training (MU-CPT), enabling deployed MLLMs to continually evolve from streaming unlabeled data. Existing unsupervised post-training methods for MLLMs typically optimize target tokens uniformly, overlooking their heterogeneous visual dependence (VD). However, we reveal that token-level VD is crucial for MU-CPT. Specifically, its structural distortion serves as an indicator of cross-modal catastrophic forgetting, and its inherent heterogeneity acts as a compass to guide new-task learning. Leveraging this property, w

Read original source ↗ Open in workspace

recordType

paper

region

Global