Rethink reporting of evaluation results in AI

Aggregate metrics and lack of access to results limit understanding.

Rethink reporting of evaluation results in AI | Litlas