5 papers
DOGMA: Weaving Structural Information into Data-centric Single-cell Transcriptomics Analysis
Ru Zhang, Xunkai Li, Yaxin Deng +7
Recently, data-centric AI methodology has been a dominant paradigm in single-cell transcriptomics analysis, which treats data representation rather than model complexity as the fun…
Toward Effective Multimodal Graph Foundation Model: A Divide-and-Conquer Based Approach
Sicheng Liu, Xunkai Li, Daohan Su +4
Graph Foundation Models (GFMs) have achieved remarkable success in generalizing across diverse domains. However, they mainly focus on Text-Attributed Graphs (TAGs), leaving Multimo…
VisuoAlign: Safety Alignment of LVLMs with Multimodal Tree Search
MingSheng Li, Guangze Zhao, Sichen Liu
Large Vision-Language Models (LVLMs) have achieved remarkable progress in multimodal perception and generation, yet their safety alignment remains a critical challenge.Existing def…
Two Facets of the Same Optimization Coin: Model Degradation and Representation Collapse in Graph Foundation Models
Xunkai Li, Daohan Su, Sicheng Liu +5
Inspired by the success of LLMs, GFMs are designed to learn the optimal embedding functions from multi-domain text-attributed graphs for the downstream cross-task generalization ca…
SITransformer: Shared Information-Guided Transformer for Extreme Multimodal Summarization
Sicheng Liu, Lintao Wang, Xiaogang Zhu +3
Extreme Multimodal Summarization with Multimodal Output (XMSMO) becomes an attractive summarization approach by integrating various types of information to create extremely concise…