AIM：多模态LLM推理中多智能体辩论的自适应交互
原标题：AIM: Adaptive Interaction in Multi-Agent Debate for Multimodal LLM Inference
AI 摘要
该论文提出AIM（自适应交互多智能体辩论）框架，用于多模态大模型推理。AIM不再固定使用文本或图结构交互，而是动态选择文本、图以及任务相关模态（图像区域、音频片段、视频时空片段）的组合来交换证据。实验覆盖VQA、Audio-QA、Video-QA共八个基准，AIM准确率最高，相比最佳单模态基线提升至多6.4%，相比各基准最优固定组合再提升至多4.2%，同时token成本降低至多53.4%。
正文节选
AIM: Adaptive Interaction in Multi-Agent Debate for Multimodal LLM Inference Abstract Multi-Agent Debate (MAD) improves reasoning on multimodal tasks, but its accuracy and token efficiency depend on how agents interact during debate. A largely overlooked design choice is which interaction modalities agents use to exchange evidence. Existing methods rely on fixed text- or graph-based interaction, which fails to convey fine-grained multimodal evidence and thus degrades accuracy while inflating to