Unverified50% confidenceFactExact time
Graph neural networks (GNN) treat crystal structures as graphs and aggregate neighbor information through message passing to predict properties such as formation energy, band gap and elastic modulus, with typical models including CGCNN, SchNet, MEGNet and ALIGNN.
1
Sources
50%
Confidence
Long-term
Relevance
10/18/2026
First Seen
Sources
Related Entities
Related Claims
Unverified混合架构的核心理念是将不同类型的神经网络模块（如RNN/GRU、Transformer注意力、GNN、统计因子分支）组合以捕捉金融数据中的多维度模式69% similarUnverifiedGNoME基于图神经网络架构，将晶体结构表示为原子节点和化学键边的图，在Materials Project等数据库中约4.8万种已知稳定材料上训练69% similarUnverified代表性等变神经网络架构包括SchNet、EGNN、SE(3)-Transformer，扩散框架相关模型包括FrameDiff、FoldFlow64% similarUnverifiedncnn is a high-performance neural network forward computation framework open-sourced by Tencent Youtu Lab, specifically optimized for mobile ARM platforms with no third-party dependencies.61% similarUnverifiedNetron is a neural network visualization tool that supports multiple model formats including ONNX, TensorFlow, and PyTorch.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1058156API
curl https://kongchang.com/api/v1/knowledge/claims/1058156MCP
get_claim(id=1058156)