List your company

Open-ended Refusal on Open-ended refusal benchmark

8.76ScoreTF
7.727.998.268.53Jun 7, 2026

Evaluation results

MethodLinks
TFModel=Qwen2.5-32B2026.068.763.85
ODEModel=Qwen2.5-32B2026.068.723.83
LAcTModel=Qwen2.5-32B2026.068.73.81
INNSTEERModel=Qwen2.5-32B2026.068.73.43
ITIModel=Qwen2.5-32B2026.068.663.82
TFModel=LLaMA3-8B2026.068.645.4
INNSTEERModel=LLaMA3-8B2026.068.624.79
INNSTEERModel=Qwen2.5-3B2026.068.625.34
INNSTEERModel=Qwen2.5-7B2026.068.625.26
TFModel=Qwen2.5-7B2026.068.65.75
ODEModel=LLaMA3-8B2026.068.65.35
INNSTEERModel=LLaMA3-3B2026.068.67.06
LAcTModel=LLaMA3-8B2026.068.585.28
ODEModel=Qwen2.5-7B2026.068.585.68
CAAModel=Qwen2.5-32B2026.068.563.78
TFModel=LLaMA3-3B2026.068.567.62
LAcTModel=Qwen2.5-7B2026.068.545.55
ODEModel=LLaMA3-3B2026.068.547.5
ITIModel=LLaMA3-8B2026.068.525.3
HPRModel=Qwen2.5-32B2026.068.523.8
ITIModel=Qwen2.5-7B2026.068.55.62
LAcTModel=LLaMA3-3B2026.068.57.48
TFModel=Qwen2.5-3B2026.068.55.82
ITIModel=LLaMA3-3B2026.068.467.55
ODEModel=Qwen2.5-3B2026.068.465.75
CAAModel=LLaMA3-8B2026.068.424.89
LAcTModel=Qwen2.5-3B2026.068.425.66
CAAModel=Qwen2.5-7B2026.068.385.66
ITIModel=Qwen2.5-3B2026.068.385.72
HPRModel=LLaMA3-8B2026.068.385.2
PoEModel=Qwen2.5-32B2026.068.363.69
CAAModel=LLaMA3-3B2026.068.347.44
HPRModel=Qwen2.5-7B2026.068.345.48
PoDModel=Qwen2.5-32B2026.068.323.68
HPRModel=LLaMA3-3B2026.068.37.3
RECModel=Qwen2.5-32B2026.068.283.67
CAAModel=Qwen2.5-3B2026.068.265.7
PoEModel=LLaMA3-8B2026.068.225.05
HPRModel=Qwen2.5-3B2026.068.225.58
BaseModel=Qwen2.5-32B2026.068.183.65
PoDModel=LLaMA3-8B2026.068.185.1
PoEModel=Qwen2.5-7B2026.068.144.69
RECModel=LLaMA3-8B2026.068.145
PoDModel=Qwen2.5-7B2026.068.14.73
PoEModel=LLaMA3-3B2026.068.16.92
RECModel=Qwen2.5-7B2026.068.064.64
PoDModel=LLaMA3-3B2026.068.046.95
BaseModel=LLaMA3-8B2026.068.025.25
PoEModel=Qwen2.5-3B2026.068.025.15
RECModel=LLaMA3-3B2026.0686.9
PoDModel=Qwen2.5-3B2026.067.985.18
RECModel=Qwen2.5-3B2026.067.945.12
BaseModel=Qwen2.5-7B2026.067.94.58
BaseModel=LLaMA3-3B2026.067.846.88
BaseModel=Qwen2.5-3B2026.067.765.1