How to contribute?

You have found a bug in TabularBench or have a suggestion for a new functionality? Then get in touch with us by opening up an issue on the github issue board to discuss possible new developments with the community and the maintainers.

Submitting a new model, dataset, attack, defense

We welcome attacks contributions that bring additional insights or challenges to the tabular adversarial robustness community.

1. Create a new issue by selecting the desired type of contribution.

2. Fill in the form accordingly.

3. We will create a new PR to implement your contributions.

4. Once ready, they will be available in the the Benchmark and if you allow it in the API.

Roadmap

• 0.1.0: Initial release w/ documentation

• Oct 2024: Distribution of code to train each tabular generator

• Jan 2025: Updated benchmark with adaptive robustness score.

• March 2025: Implementation of additional distances