Knowledge Editing
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
91.6H-Acc
18
May 28, 2026
89.8H-Acc
18
May 28, 2026
0.65Normalized Editing Time
18
Feb 26, 2026
96.6Bert Score
18
Feb 26, 2026
99.86BERTScore
18
Feb 26, 2026
99.95Accuracy
17
May 4, 2026
50.417Overall M-Acc.
16
Apr 8, 2026
52.17Precision
16
Feb 26, 2026
90.17Reliability
16
Feb 26, 2026
100Reliability
16
Feb 26, 2026
1Reliability
16
Feb 26, 2026
1Reliability
16
Feb 26, 2026
100Reliability
16
Feb 26, 2026
88.57Average EM
15
Jun 17, 2026
100Metric T
15
Jun 17, 2026
99.38T-EM
15
Jun 17, 2026
99.72Target-based EM
15
Jun 17, 2026
100Reliability (Rel)
14
May 29, 2026
85.4H-Acc
14
May 28, 2026
91.3H-Acc
14
May 28, 2026
100Fact Accuracy (Easy)
14
Feb 26, 2026
100Fact Accuracy (Easy)
14
Feb 26, 2026
99.9Fact Accuracy (Easy)
14
Feb 26, 2026
92.8Math BLEU
12
Jun 2, 2026
1Reliability
12
May 12, 2026