Instruction Following
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
76.5Accuracy
41
Jul 7, 2026
85.56Score
41
Jun 9, 2026
47.24ROUGE Score
40
Mar 4, 2026
99.88P. Score
40
Mar 4, 2026
387.196Reward Score
40
Feb 26, 2026
67.1Accuracy
40
Apr 21, 2026
81.2Win Rate
36
Jun 2, 2026
46.6Self-BLEU
36
May 4, 2026
84.7Score
36
Apr 6, 2026
36.86ROUGE-L
36
Mar 4, 2026
36.82ROUGE-L
36
Mar 4, 2026
20.37ROUGE-L
36
Mar 4, 2026
55.95Score
36
Jun 30, 2026
62RougeL
36
Feb 26, 2026
71.3Score
36
Jun 9, 2026
8.33Overall Score
35
May 19, 2026
83.7Inv-IF
34
Jul 2, 2026
92.2Win Rate
34
May 7, 2026
73.9Score
34
May 29, 2026
31.1ROUGE-L
33
Mar 5, 2026
25.39Self-Instruct Score
33
Feb 26, 2026
79.81F1 Score
32
May 13, 2026
82Win Rate
32
Jun 4, 2026
94.6ISR
32
Feb 26, 2026
4.5Wrong Rate (%)
31
Jun 11, 2026