1 paper · 1 filter
Abhilash Mishra, Yash Gorana
Labelled "ground truth" datasets are routinely used to evaluate and audit AI algorithms applied in high-stakes settings. However, there do not exist widely accepted benchmarks for…