Instruction Following
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
62A Win Count
19
Feb 26, 2026
56.7Win Rate (A)
19
Feb 26, 2026
66.3Win Rate (A)
19
Feb 26, 2026
90Score
18
Jun 23, 2026
90.58Loose Score (MaXIFE)
18
May 22, 2026
96.3IFEval Score
18
May 22, 2026
36Accuracy
18
May 19, 2026
63.18WB Score
18
Apr 16, 2026
73.5Example Score
18
Apr 16, 2026
78.1Avg@8
18
May 12, 2026
93.89IF Rate
18
May 21, 2026
52IFEval Score
18
Mar 12, 2026
84.9Win Rate
18
May 7, 2026
12.09AlpacaEval Score
18
Feb 26, 2026
8.09GPT-4 Evaluation Score
18
Feb 26, 2026
31.3AlpacaEval Win Rate (GPT-4)
18
Apr 29, 2026
62.47ROUGE-L Score
18
Feb 26, 2026
53.53ROUGE-L
18
Feb 26, 2026
100Accuracy
18
Feb 26, 2026
58.2Score
18
Feb 26, 2026
88.36Accuracy
17
Jul 8, 2026
83.29Overall CSR
17
Jun 9, 2026
57.7Score
17
May 29, 2026
55.79Instruction Following Score (ISR)
17
Apr 21, 2026
57.8ISR
17
Apr 21, 2026