Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Instruction Following on ComplexBench

85.22Overall Score

ImpRIF-32B

-2.84792820.01586142.8796565.743439Feb 17, 2025May 8, 2025Jul 27, 2025Oct 15, 2025Jan 3, 2026Mar 24, 2026Jun 13, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
85.22-
2026.02
84.1-
2026.02
83.46-
2026.02
83.39-
2026.02
83.29-
2026.02
82.16-
2026.02
81.99-
2026.02
81.56-
2026.02
81.48-
2026.02
81.37-
2026.02
81.22-
2026.02
80.91-
2026.02
79.37-
2026.02
77.57-
2026.02
77.04-
2026.02
76.8-
2026.02
73.76-
2026.06
71.686.7
59.685.6
56.287.6
55.487.4
2026.06
54.883.9
2026.06
54.283.1
2026.06
51.283.5
2026.06
50.486.1
2026.06
47.985
2026.06
45.179.2
2026.06
43.679.8
2026.06
40.677
2025.02
0.7-
2025.02
0.6939-
2025.02
0.6743-
2025.02
0.6732-
2025.02
0.6724-
2025.02
0.6713-
2025.02
0.6702-
2025.02
0.6698-
2025.02
0.6645-
2025.02
0.6598-
2025.02
0.6589-
2025.02
0.6473-
2025.02
0.6441-
2025.02
0.6432-
2025.02
0.643-
2025.02
0.6413-
2025.02
0.6405-
2025.02
0.6352-
2025.02
0.6292-
2025.02
0.6267-
2025.02
0.6245-
2025.02
0.6203-
2025.02
0.6149-
2025.02
0.6079-
2025.02
0.5591-
2025.02
0.5393-