List your company

Refusal on Refusal benchmark

97.81Alignment ProbabilityINNSTEER
32.383649.369366.35583.3407Jun 7, 2026

Evaluation results

MethodLinks
INNSTEERBase Model=LLaMA3-3B2026.0697.815.54
INNSTEERBase Model=LLaMA3-8B2026.0691.93.63
INNSTEERBase Model=Qwen2.5-7B2026.0686.462.92
ODEBase Model=LLaMA3-3B2026.0684.65.2
TFBase Model=LLaMA3-3B2026.0682.45.45
ITIBase Model=LLaMA3-3B2026.0682.15.35
LAcTBase Model=LLaMA3-3B2026.0681.55.3
INNSTEERBase Model=Qwen2.5-32B2026.0681.093.65
HPRBase Model=LLaMA3-3B2026.0680.755.1
CAABase Model=LLaMA3-3B2026.0678.475.24
RECBase Model=LLaMA3-3B2026.0676.64.9
BaseBase Model=LLaMA3-3B2026.0676.524.89
PoEBase Model=LLaMA3-3B2026.0675.254.93
PoDBase Model=LLaMA3-3B2026.0675.14.95
ODEBase Model=LLaMA3-8B2026.0673.53.27
ODEBase Model=Qwen2.5-32B2026.0672.33.7
TFBase Model=LLaMA3-8B2026.0671.23.32
ITIBase Model=LLaMA3-8B2026.0670.43.3
LAcTBase Model=LLaMA3-8B2026.0669.93.29
TFBase Model=Qwen2.5-32B2026.0669.53.42
HPRBase Model=LLaMA3-8B2026.0669.13.28
ITIBase Model=Qwen2.5-32B2026.0668.33.1
LAcTBase Model=Qwen2.5-32B2026.0667.83.09
HPRBase Model=Qwen2.5-32B2026.0666.93.08
ODEBase Model=Qwen2.5-7B2026.0666.12.86
INNSTEERBase Model=Qwen2.5-3B2026.0665.122.92
CAABase Model=LLaMA3-8B2026.0665.113.16
RECBase Model=LLaMA3-8B2026.0664.83.22
BaseBase Model=LLaMA3-8B2026.0664.683.21
PoEBase Model=LLaMA3-8B2026.0663.353.24
PoDBase Model=LLaMA3-8B2026.0663.23.25
TFBase Model=Qwen2.5-7B2026.0663.22.89
ITIBase Model=Qwen2.5-7B2026.0662.52.88
LAcTBase Model=Qwen2.5-7B2026.0661.82.87
CAABase Model=Qwen2.5-32B2026.0661.452.96
HPRBase Model=Qwen2.5-7B2026.0660.92.86
RECBase Model=Qwen2.5-32B2026.0659.63.02
BaseBase Model=Qwen2.5-32B2026.0659.183
PoEBase Model=Qwen2.5-32B2026.0658.33.04
PoDBase Model=Qwen2.5-32B2026.0658.13.05
CAABase Model=Qwen2.5-7B2026.0656.22.81
RECBase Model=Qwen2.5-7B2026.0653.52.83
BaseBase Model=Qwen2.5-7B2026.0653.112.82
PoEBase Model=Qwen2.5-7B2026.06522.84
PoDBase Model=Qwen2.5-7B2026.0651.82.85
ODEBase Model=Qwen2.5-3B2026.0648.83.18
TFBase Model=Qwen2.5-3B2026.0646.13.22
ITIBase Model=Qwen2.5-3B2026.0645.23.2
LAcTBase Model=Qwen2.5-3B2026.0644.73.19
HPRBase Model=Qwen2.5-3B2026.0643.83.18
CAABase Model=Qwen2.5-3B2026.0638.583.11
RECBase Model=Qwen2.5-3B2026.0636.23.12
BaseBase Model=Qwen2.5-3B2026.0635.783.11
PoEBase Model=Qwen2.5-3B2026.0635.053.14
PoDBase Model=Qwen2.5-3B2026.0634.93.15