Paraphrase Identification
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
92AUROC (QQP->QQP)
10
Feb 26, 2026
0.849AUROC
10
Feb 26, 2026
0.85AUROC
10
Feb 26, 2026
85.1AUROC
10
Feb 26, 2026
73Macro F1
10
Feb 26, 2026
81Macro F1
10
Feb 26, 2026
84.7Accuracy (de)
7
Feb 27, 2026
70.6Accuracy
7
Feb 26, 2026
63.99AUROC
5
Jun 5, 2026
70.88Accuracy (Clean)
5
Apr 28, 2026
76Accuracy
5
Feb 26, 2026
20.44ARG (NP -> P)
5
Feb 26, 2026
90.14Accuracy
5
Feb 26, 2026
87.73F1 Score
5
Feb 26, 2026
83.03Accuracy
5
Feb 26, 2026
22.69Latency (ms)
4
May 12, 2026
55.5Accuracy (en)
2
Feb 27, 2026
78.7Max F1 Score
2
Feb 26, 2026
0.755F1 Score
2
Feb 26, 2026
88.9F1 Score
2
Feb 26, 2026