Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
2
0
4
By keeping benchmark-native outcomes distinct from pooled summaries, the study makes test-population, class-balance, and missing-record-coverage differences visible and supports interpretation of detector results in research, platform-safety, and forensic-review settings, foregrounding traceable protocol conditions over claims or leaderboard comparisons.
A striking divergence between ranking ability and operating-point quality reveals that high ROC-AUC scores can mislead deepfake detection effectiveness.