ResearchBenchmarksDiscrimination between Good Faith and Problematic agents on WMT14 1.1:1Follow3.32Cohen's dTVD-MI0.11680.94841.782.6116Aug 7, 2025Evaluation ResultsMethodMethodLinksCohen's dTVD-MI2025.083.32JudgeReference availability...Reference availability=ref2025.082.53MI (DoE)2025.081.61Baseline2025.080.93GPPM2025.080.7JudgeReference availability...Reference availability=no ref2025.080.24