weirayao / leap
LEAP is a tool for discovering latent temporal causal relations with gradient-based neural network.
☆35Updated 2 years ago
Alternatives and similar repositories for leap:
Users that are interested in leap are comparing it to the libraries listed below
- Disentangled gEnerative cAusal Representation (DEAR)☆60Updated 2 years ago
- ☆24Updated 3 weeks ago
- Code for "Causal autoregressive flows" - AISTATS, 2021☆45Updated 4 years ago
- On the Role of Sparsity and DAG Constraints for Learning Linear DAGs☆33Updated 3 years ago
- Code for the paper "Disentanglement by Nonlinear ICA with General Incompressible-flow Networks (GIN)" (2020)☆32Updated 3 years ago
- Continuous-Time Modeling of Counterfactual Outcomes Using Neural Controlled Differential Equations (ICML 2022)☆27Updated 2 years ago
- ☆25Updated 3 years ago
- VAEs and nonlinear ICA: a unifying framework