Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Deep Structural Causal Models for Tractable Counterfactual Inference
Authors: Nick Pawlowski, Daniel Coelho de Castro, Ben Glocker
NeurIPS 2020 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our framework is validated on a synthetic dataset built on MNIST as well as on a real-world medical dataset of brain MRI scans. Our experimental results indicate that we can successfully train deep SCMs that are capable of all three levels of Pearl s ladder of causation: association, intervention, and counterfactuals, giving rise to a powerful new approach for answering causal questions in imaging applications and beyond. |
| Researcher Affiliation | Academia | Nick Pawlowski Imperial College London EMAIL Daniel C. Castro Imperial College London EMAIL Ben Glocker Imperial College London EMAIL |
| Pseudocode | No | The paper describes methods and processes in narrative text and figures, but it does not include any explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | The code for all our experiments is available at https://github.com/biomedia-mira/deepscm. |
| Open Datasets | Yes | Our framework is validated on a synthetic dataset built on MNIST as well as on a real-world medical dataset of brain MRI scans. ... Modelling brain MRI scans ... using population data from the UK Biobank [54]. |
| Dataset Splits | Yes | We used 80% of the data for training, and 20% for validation/testing. |
| Hardware Specification | Yes | We gratefully acknowledge the support of NVIDIA with the donation of one Titan X GPU. |
| Software Dependencies | Yes | All models were implemented using Py Torch v1.6.0 and Pyro v1.4.1. |
| Experiment Setup | Yes | A total of 100,000 samples were used for training. Adam [58] was used as optimizer with a learning rate of 10−4. All models were trained for 100 epochs, with Adam [58] optimizer and learning rate of 10−4. We used 80% of the data for training, and 20% for validation/testing. |