Instruction Following
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
55.14Original Ratio
2
Jun 18, 2026
53.88Original Ratio
2
Jun 18, 2026
66.66MMLU Score
2
Jun 9, 2026
17.93Instruction Following (Loose)
2
May 27, 2026
11.4Avg Steps per Run
2
May 15, 2026
99.48F1 Score
2
May 13, 2026
67.8RM Score
2
May 11, 2026
84.7Score
2
May 1, 2026
90.2IFR (%)
2
Mar 12, 2026
71WinRate
2
Mar 12, 2026
96IFEval Score
2
Mar 12, 2026
85.9Accuracy
2
Mar 12, 2026
446Pass@1
2
Feb 26, 2026
0.755Paraphrase
2
Feb 26, 2026
1.61Speedup
2
Feb 26, 2026
31.5Avg@10
2
Feb 26, 2026
80.47Overall Score
2
Feb 26, 2026
5.96ChatGPT-Rank
2
Feb 26, 2026
8.1Helpfulness
2
Feb 26, 2026
70.1Instruction Following Rate
2
Feb 26, 2026
96.9Instruction Following Rate
2
Feb 26, 2026
95.9Instruction Following Rate
2
Feb 26, 2026
22.5Win-Loss Differential (%)
2
Feb 26, 2026
15Win-Loss Differential (%)
2
Feb 26, 2026
53.8Win-Loss Differential (%)
2
Feb 26, 2026