CHEM-AD: MOFxDB descriptors and anomaly scores (v1.0)
收藏官方服务：
资源简介：
We release the data underlying CHEM-AD, a CPU-efficient, label-free anomaly detector for metal–organic frameworks (MOFs) trained on 32 standardized descriptors.Scope. 26,025 MOFxDB entries; anomaly scores from a 32–16–8–16–32 autoencoder (1,352 params) trained with MSE; knee threshold τ = 0.4666 flags 729 anomalies (≈2.8%).
提供机构：
Zenodo创建时间：
2025-08-18
