Generation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
90.68F1 Score
58
Jun 5, 2026
48.74F1 Score
49
Jun 5, 2026
4.14Average Wall-Clock Time (s)
40
May 26, 2026
2.47Average Generation Time (s)
40
May 26, 2026
2.86Average Generation Time (s)
40
May 26, 2026
64.5Rouge Score
37
Jun 2, 2026
69.03Accuracy
32
Mar 6, 2026
57.87Accuracy
32
Mar 6, 2026
67.33Accuracy
32
Mar 6, 2026
1ASR-G Similarity (mean) @-1
23
Apr 1, 2026
57.9Exact Match
17
Feb 26, 2026
86.5Perception
15
Mar 27, 2026
0.644VPTT Score (Novelty Adjusted)
15
Feb 26, 2026
81.6BERT-Score
12
May 1, 2026
0.008Data Metric Value
10
May 21, 2026
1.19Data Fidelity Deviation
10
May 21, 2026
93ASR-G Sim^-1 (Mean)
10
Apr 1, 2026
37.3Ans-F1
8
Jun 9, 2026
58.43-Class Accuracy
8
Jun 9, 2026
83.63-Class Accuracy
8
Jun 9, 2026
40.279FID (5k)
8
Apr 9, 2026
79.673FID@5k
8
Apr 9, 2026
0.07MMD
7
Mar 12, 2026
0.05MMD
7
Mar 12, 2026
0.04MMD
7
Mar 12, 2026