Welcome to the ARC (AI Rating through Causality) Tool!

ARC evaluates AI systems for bias and robustness using causal analysis, encompassing tasks such as binary classification, sentiment analysis, group recommendation, and time-series forecasting. The framework is designed to seamlessly extend to other tasks. Choose task, data, attributes, test systems, and metrics from a comprehensive suite of pre-existing options to get in-depth causal analysis results that quantify the bias or robustness exhibited by the systems. Click on 'Proceed' to get started!

Here is our rating workflow for performing statistical and causal analysis to compute raw scores and assign final ratings to the test systems. For more information, please visit our project page. The demonstration video provides a guided walkthrough of the tool’s background and functionality.