Digital circuits dataset for anomaly detection + ground truth explanations
收藏NIAID Data Ecosystem2026-03-12 收录
下载链接:
https://doi.org/10.7910/DVN/W4FPPN
下载链接
链接失效反馈官方服务:
资源简介:
This dataset is an anomaly benchmark dataset with local ground-truth explanations. It was created based on 4 digital circuits from ISCAS '85 and 74x series benchmarks: C17, 74182, 74283 and 74181. The dataset includes: Truth_Tables Each circuits has 4 anomalous versions created by changing a single gate to its negation. Each anomalous version was padded with a varying amount of attribute noise (2, 4 and 6 noise features). The folder includes a truth table for each version , original and anomalous with and without noise. Overall there are 64 anomalous truth tables - named "in_labeled", and 16 original truth table- named "out". Ground_Truth Each truth table has a file of local ground-truth explanations. A local ground truth explanation is the reason why a model returned a certain prediction for a specific instance. Such an explanation may be represented as the set of features that led the model to make that prediction. Both the Truth_Tabels and Ground_Truth tables are divided into circuits folders.
创建时间:
2021-09-05



