Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Hallucination Detection on SciQ (AUROC, F1-Score)

0.9328AUROC

CausalGaze

0.5076480.6180240.72840.838776Apr 13, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
0.932882.83
2026.04
0.921784.02
2026.04
0.85578.6
2026.04
0.83977.01
2026.04
0.834372.73
2026.04
0.833676.49
2026.04
0.81582
2026.04
0.788477.9
2026.04
0.779378.26
2026.04
0.774875.61
2026.04
0.77377.67
2026.04
0.759572.15
2026.04
0.694566.4
2026.04
0.68267.97
2026.04
0.67968.3
2026.04
0.65966.04
2026.04
0.647463.46
2026.04
0.63264.62
2026.04
0.603658.32
2026.04
0.59660.08
2026.04
0.58657.79
2026.04
0.584256.37
2026.04
0.57658.26
2026.04
0.574855.47
2026.04
0.572657.73
2026.04
0.567355.31
2026.04
0.559352.6
2026.04
0.54755.83
2026.04
0.54656.45
2026.04
0.53854.49
2026.04
0.53454.77
2026.04
0.52652.56
2026.04
0.52452.56