M³Prune：面向高效多模态多智能体检索增强生成的分层协同剪枝
M$^3$Prune: Hierarchical Collaborative Pruning for Efficient Multi-Modal Multi-Agent Retrieval-Augmented Generation
浏览论文内容
中文总结 AI 辅助
M³Prune是优化多模态多智能体检索增强生成的分层协同剪枝框架，通过剪枝冗余通信边提升性能与token效率，实验中优于单智能体及多智能体基准系统。
中文摘要 AI 辅助
近期多模态检索增强生成（mRAG）领域的进展，即利用外部知识增强多模态大语言模型（MLLM），已表明多智能体的集体智能通过有效通信可优于单一模型。尽管现有多智能体系统性能强劲，但会产生大量token开销与计算成本，对大规模部署构成挑战。为解决这些问题，我们提出多模态多智能体分层通信图剪枝框架，命名为M³Prune。M³Prune消除跨模态及模态内部的冗余通信边，优化任务性能与token开销的权衡。具体而言，M³Prune先在文本与视觉模态内执行图稀疏化，识别任务关键通信链路；再构建跨模态通信图并稀疏化跨模态连接，同时通过模态对齐分数鼓励一致的跨模态推理；最后逐步剪枝冗余边，获得高效分层拓扑。在通用领域及特定领域mRAG基准上的大量实验表明，M³Prune始终优于单智能体及强大的多智能体mRAG系统，且显著提升token效率。
英文摘要
Recent advances in multi-modal retrieval-augmented generation (mRAG), which augments multi-modal large language models (MLLMs) with external knowledge, have shown that collective intelligence from multiple agents can outperform a single model through effective communication. Despite their strong performance, existing multi-agent systems incur substantial token overhead and computational cost, posing challenges for large-scale deployment. To address these issues, we propose a Multi-Modal Multi-agent hierarchical communication graph PRUNING framework, termed M3Prune. M3Prune eliminates redundant communication edges both across and within modalities, improving the trade-off between task performance and token overhead. Specifically, M3Prune first performs intra-modal graph sparsification in the textual and visual modalities to identify task-critical communication links. It then constructs an inter-modal communication graph and sparsifies cross-modal connections while encouraging consistent cross-modal reasoning through a modality alignment score. Finally, it progressively prunes redundant edges to obtain an efficient hierarchical topology. Extensive experiments on both general-domain and domain-specific mRAG benchmarks show that M3Prune consistently outperforms single-agent and strong multi-agent mRAG systems while signifi- cantly improving token efficiency.
