medical
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Medical
68.81GPT Accuracy
31
Medical (Med)
0.1PPL Change (%) vs Baseline
30
medical
87.5Average Precision
21
medical
0.03Ranking Loss
21
Medical
50.9Claude-4 Sonnet Score
21
medical Gold standard domain-specific (test)
77.32MRR
18
Medical
100Accuracy
16
Medical
0.871PLCC
15
Medical Random subset
25.04R-LCS
14
Medical
81.55Score
14
Medical
8.58Avg Score
14
Medical Retina (test)
39.76AUC @ 3px Tolerance
13
Medical Average Global
1Spearman Correlation
12
Medical Ru→En
100HitRate@p
12
Medical Zh→En
100HitRate@p
12
Medical En→Ru
100Hit Rate@p
12
Medical En→Zh
100HitRate@p
12
Medical
12.6Attack Success Rate (Atk)
11
Medical
85.1Attack Accuracy
11
Medical
7Indexing Time (minutes)
11
Medical
76.6Micro F1-Score
11
Medical
76.6F1 Score
10
Medical
36.2Leakage (Deg)
10
Medical
87.3SAFE
10
Medical (test)
55.42BLEU
9