List your company

Jailbreak Evaluation on WildJailbreak

98.5Performance RateLLM w/o cascade
14.7836.51558.2579.985May 17, 2026

Evaluation results

MethodLinks
LLM w/o cascadeCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE2026.0598.540.043.49100100——
LLM w/o cascadeCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B2026.0598.540.043.49——100—
InitialCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE2026.059820.732.115834——
InitialCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B2026.059724.82.39——67—
Joint-target attackCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE, Setting=P=12026.0588.539.173.0488.537——
Joint-target attackCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B, Setting=P=12026.0586.537.443.18——91.5—
Random noiseCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B2026.0584.528.972.54——77.5—
Random noiseCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE2026.058223.852.2461.511.5——
Single-target attackCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B, Setting=SLM only2026.058129.632.67——78.5—
Single-target attackCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE, Setting=SLM only2026.057930.122.5561.537.5——
Joint-target attackCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE, Setting=P=0.82026.0574.532.042.627433.5——
Joint-target attackCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B, Setting=P=0.82026.057330.982.71——79—
Single-target attackCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B, Setting=DM only2026.056619.272.16——53—
Single-target attackCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE, Setting=DM only2026.056417.81.9446.514——
Joint-target attackCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE, Setting=P=0.62026.0561.526.552.446627——
Joint-target attackCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B, Setting=P=0.62026.0558.525.122.44——67.5—
Joint-target attackCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE, Setting=P=0.42026.055020.941.945430.5——
Joint-target attackCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B, Setting=P=0.42026.0547.519.892.08——54.5—
Joint-target attackCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE, Setting=P=0.22026.0541.513.81.6843.520.5——
Joint-target attackCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B, Setting=P=0.22026.0538.513.71.69——38.5—
Joint-target attackCascade system=Qwen2.5 + BERT-base + Mistral-7B + BERT-base + Phi-3.5-MoE, Setting=P=02026.05317.911.2228.516——
Joint-target attackCascade System=Qwen2.5-3B + DistilBERT-base + Phi-3.5-MoE 16x3.8B, Setting=P=02026.05186.431.11——22.5—
BaseBackbone=Qwen3-4B2026.06——————69.46
ICLBackbone=Qwen3-4B2026.06——————21.09
PolicyAlignBackbone=Qwen3-4B2026.06——————8.51
SFTBackbone=Qwen3-4B2026.06——————9.73