captum
Interpretability for sequence generation models 🐛 🔍
Collection of NLP model explanations and accompanying analysis tools
Build sparsely connected PyTorch neural networks from prior-knowledge graphs, with optional feature- and node-level attribution.
