2025/06/30 by Dai, Yang, James An, An, Jianxiang +12 · 1 citation
Computer Science · #Advanced Graph Neural Networks #Computation and Language (cs.CL) #Data integration #Domain (mathematical analysis) #Domain Adaptation and Few-Shot Learning #Domain knowledge #Exploit #FOS: Computer and information sciences #Inference #Knowledge base #Modular design #Scalability #Task (project management) #Topic Modeling
paper · pdf · doi:10.48550/arxiv.2506.23940
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2025/06/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
Multimodal Large Language Models (MLLMs) have achieved success across various domains. However, their applicability tends to degrade when confronted with different types of data inputs, especially for MLLMs that have been fine-tuned for specific tasks. Despite its importance, the study of knowledge sharing among domain-specific MLLMs--such as those trained for mathematics or code--remains largely underexplored. To address the fragmentation of knowledge across domain-specialized MLLMs, we propose a unified parameter integration framework that enables modular composition of expert capabilities. Our method is grounded in a novel Compatibility-Aware Parameter Splicing (CAPS) strategy, which leverages both local functional attribution and global information-theoretic signals to guide selective parameter fusion. By extending this mechanism to the low-rank adaptation layer granularity, we ensure efficient integration with minimal inference overhead. Furthermore, we introduce a domain compatibility scoring mechanism that quantifies inter-expert alignment at the activation level and correlates with downstream task utility. This principled fusion protocol allows the final model to synergize heterogeneous expertise while preserving structural modularity. Extensive evaluations across diverse multimodal benchmarks validate the effectiveness of our framework, offering a scalable path toward compositional, domain-adaptive MLLMs.