List your company

Safety Evaluation on BeaverTails & DirectRefusal Average

32.84Average ASRSafeGene
31.964437.874743.78549.6953Jun 2, 2026

Evaluation results

MethodLinks
SafeGeneBackbone=Qwen2.5-7B, Downstream Task Context=MNLI, Safety Judge=Beaver-dam-7B2026.0632.84
SafeGeneBackbone=Qwen2.5-7B, Downstream Task Context=BoolQ, Safety Judge=Beaver-dam-7B2026.0633
SafeGeneBackbone=Qwen2.5-7B, Downstream Task Context=AG News, Safety Judge=Beaver-dam-7B2026.0633.45
SafeGeneBackbone=Qwen2.5-7B, Downstream Task Context=Average, Safety Judge=Beaver-dam-7B2026.0633.84
SaLoRABackbone=Qwen2.5-7B, Downstream Task Context=BoolQ, Safety Judge=Beaver-dam-7B2026.0635.54
SafeGeneBackbone=Qwen2.5-7B, Downstream Task Context=SST2, Safety Judge=Beaver-dam-7B2026.0636.05
SaLoRABackbone=Qwen2.5-7B, Downstream Task Context=AG News, Safety Judge=Beaver-dam-7B2026.0636.64
SaLoRABackbone=Qwen2.5-7B, Downstream Task Context=Average, Safety Judge=Beaver-dam-7B2026.0636.84
SafeMERGEBackbone=Qwen2.5-7B, Downstream Task Context=MNLI, Safety Judge=Beaver-dam-7B2026.0636.89
SaLoRABackbone=Qwen2.5-7B, Downstream Task Context=SST2, Safety Judge=Beaver-dam-7B2026.0636.95
SafeMERGEBackbone=Qwen2.5-7B, Downstream Task Context=BoolQ, Safety Judge=Beaver-dam-7B2026.0637.12
SafeMERGEBackbone=Qwen2.5-7B, Downstream Task Context=Average, Safety Judge=Beaver-dam-7B2026.0637.75
SafeMERGEBackbone=Qwen2.5-7B, Downstream Task Context=SST2, Safety Judge=Beaver-dam-7B2026.0638.02
SaLoRABackbone=Qwen2.5-7B, Downstream Task Context=MNLI, Safety Judge=Beaver-dam-7B2026.0638.23
SafeMERGEBackbone=Qwen2.5-7B, Downstream Task Context=AG News, Safety Judge=Beaver-dam-7B2026.0638.97
SafeLoRABackbone=Qwen2.5-7B, Downstream Task Context=BoolQ, Safety Judge=Beaver-dam-7B2026.0639.64
SafeLoRABackbone=Qwen2.5-7B, Downstream Task Context=Average, Safety Judge=Beaver-dam-7B2026.0640.63
Fine-tunedBackbone=Qwen2.5-7B, Downstream Task Context=MNLI, Safety Judge=Beaver-dam-7B2026.0640.71
SafeLoRABackbone=Qwen2.5-7B, Downstream Task Context=AG News, Safety Judge=Beaver-dam-7B2026.0640.87
SafeLoRABackbone=Qwen2.5-7B, Downstream Task Context=MNLI, Safety Judge=Beaver-dam-7B2026.0640.89
SafeLoRABackbone=Qwen2.5-7B, Downstream Task Context=SST2, Safety Judge=Beaver-dam-7B2026.0641.14
Fine-tunedBackbone=Qwen2.5-7B, Downstream Task Context=BoolQ, Safety Judge=Beaver-dam-7B2026.0646.38
Fine-tunedBackbone=Qwen2.5-7B, Downstream Task Context=Average, Safety Judge=Beaver-dam-7B2026.0647.46
Fine-tunedBackbone=Qwen2.5-7B, Downstream Task Context=AG News, Safety Judge=Beaver-dam-7B2026.0648
Fine-tunedBackbone=Qwen2.5-7B, Downstream Task Context=SST2, Safety Judge=Beaver-dam-7B2026.0654.73