Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Instruction Hierarchy Adherence on IHEval Aligned Multi-turn
Loading...
88.8
Accuracy
Prompting
33.888
48.144
62.4
76.656
Jun 29, 2026
Accuracy
Updated 26d ago
Evaluation Results
Method
Method
Links
Accuracy
Prompting
Model=Qwen3-8B-Instruct
2026.06
88.8
IHDec + VerIH
Model=Qwen3-8B-Instruct
2026.06
87.3
VerIH
Model=Qwen3-8B-Instruct
2026.06
86.29
Vanilla
Model=Qwen3-8B-Instruct
2026.06
85.6
IHDec
Model=Qwen3-8B-Instruct
2026.06
83.5
IHDec + Prompting
Model=Qwen3-4B-Instruct
2026.06
82.2
Prompting
Model=Qwen3-4B-Instruct
2026.06
78.6
IHDec + Prompting
Model=Qwen3-8B-Instruct
2026.06
77.4
IHDec
Model=Qwen3-4B-Instruct
2026.06
76.7
Vanilla
Model=Qwen3-4B-Instruct
2026.06
75.1
IHDec
Model=Llama-3.1-8B-Ins...
2026.06
68.38
Vanilla
Model=Llama-3.1-8B-Ins...
2026.06
65.9
IHDec + Prompting
Model=Llama-3.1-8B-Ins...
2026.06
57.3
Prompting
Model=Llama-3.1-8B-Ins...
2026.06
53.9
IHDec + ISE
Model=Llama-3.1-8B-Ins...
2026.06
37.6
ISE
Model=Llama-3.1-8B-Ins...
2026.06
36
Feedback
Search any
task
Search any
task