arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2406.13265cs.LGcond-mat.mtrl-sci
收藏

具有多体等变相互作用的分子图网络

Molecule Graph Networks with Many-body Equivariant Interactions

Graduate School of Frontier Sciences, The University of Tokyo(东京大学前沿科学研究生院);School of Physical and Mathematical Sciences, Nanyang Technological University(南洋理工大学物理与数学科学学院);Center for Basic Research on Materials, National Institute for Materials Science(材料基础研究机构,日本国家材料科学研究所);RIKEN Center for Advanced Intelligence Project(RIKEN高级智能项目中心)

Zetian Mao, Chuan-Shen Hu, Jiawen Li, Chen Liang, Diptesh Das, Masato Sumita, Kelin Xia, Koji Tsuda

AI总结:

针对消息传递神经网络中两体键向量抵消导致方向信息丢失的问题,提出显式整合l=1等变多体相互作用的ENINet,经数学证明必要性并推广到N体,实验显示其提升了多种量子化学性质的预测精度。

摘要:

消息传递神经网络在预测分子相互作用方面已展现出显著效能。引入等变向量表示通过捕捉几何数据对称性提升了表达能力,进而提高模型精度。然而,反向的两体键向量在消息传递过程中可能相互抵消,导致其共享节点的方向信息丢失。本研究开发了等变N体相互作用网络(Equivariant N-body Interaction Networks,ENINet),该网络显式整合l=1等变多体相互作用,以增强消息传递机制中的方向对称信息。我们通过数学分析证明了纳入多体等变相互作用的必要性,并将公式推广至N体相互作用。实验表明,整合多体等变表示可提升多种标量和张量量子化学性质的预测精度。

英文摘要:

Message passing neural networks have demonstrated significant efficacy in predicting molecular interactions. Introducing equivariant vectorial representations augments expressivity by capturing geometric data symmetries, thereby improving model accuracy. However, two-body bond vectors in opposition may cancel each other out during message passing, leading to the loss of directional information on their shared node. In this study, we develop Equivariant N-body Interaction Networks (ENINet) that explicitly integrates l = 1 equivariant many-body interactions to enhance directional symmetric information in the message passing scheme. We provided a mathematical analysis demonstrating the necessity of incorporating many-body equivariant interactions and generalized the formulation to $N$-body interactions. Experiments indicate that integrating many-body equivariant representations enhances prediction accuracy across diverse scalar and tensorial quantum chemical properties.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑