A multi-model evaluation of RECEIPT's usefulness assessment with Constitutional AI self-critique.
Run the eval to generate a multi-model quality scoring report with Constitutional AI self-critique.