发表机构
The Hong Kong University of Science and Technology (Guangzhou); Southeast University; Fudan University（香港科技大学（广州）; 东南大学; 复旦大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对多光谱序列直接连接的性能下降问题，提出MM-Spectrum稳定MoE框架，引入模态感知路由与异质专家，在分子结构解析的多模态设置下取得显著改进。
AI 中文摘要
从多模态光谱测量中推断分子结构需要整合互补但高度异质的信号。然而，直接连接多光谱序列的常见范式会出现异常的性能下降，主要原因是各模态间存在显著的异质性和由此产生的多模态不平衡。作为解决方案，我们提出MM-Spectrum，一种专为多模态多光谱谱-结构解析设计的稀疏混合专家（Mixture-of-Experts，MoE）框架。为更好匹配多光谱不平衡下的信息特征，MM-Spectrum引入显式模态感知路由机制，除令牌内容表示外，还将光谱身份暴露给路由模块。此外，它结合共享专家、交互专家及异质专家容量，以提取多模态光谱的模态独有信息和跨模态协同信息，同时抑制噪声诱导的干扰。在分子结构解析的全模态、双模态及缺失模态设置下，MM-Spectrum实现了持续且显著的改进，并有 ablation 研究和可解释性分析作为支撑。
英文摘要
Inferring molecular structures from multimodal spectroscopic measurements requires integrating complementary yet highly heterogeneous signals. However, the common paradigm of directly concatenating multispectral sequences can exhibit anomalous performance degradation, primarily due to pronounced heterogeneity and the resulting multimodal imbalance across modalities. As a remedy, we propose MM-Spectrum, a sparse Mixture-of-Experts framework tailored for multimodal multispectral spectra-to-structure elucidation. To better match the information characteristics under multispectral imbalance, MM-Spectrum introduces an explicit modality-aware routing mechanism that exposes spectral identity to the router in addition to token content representations. Moreover, it incorporates shared and interaction experts, together with heterogeneous expert capacities, to extract multispectral modality-unique and cross-modal synergistic information while suppressing noise-induced interference. Across full-modality, bimodal, and missing-modality settings on molecular structural elucidation, MM-Spectrum achieves consistent and substantial improvements, supported by ablation studies and interpretability analyses.
CommentsAccepted to ICML 2026
