List your company

Security Interception on Bilingual Security Benchmark English

98.5IRQwen2.5-14B
-3.9422.65549.2575.845Mar 7, 2026

Evaluation results

MethodLinks
Qwen2.5-14BCondition=Qwen2.5-14B2026.0398.520.188.210098.397.188.3
Llama-3.1-8BCondition=Llama-3.1-8B2026.0397.837.776.198.697.297.880.8
GPT-4o-miniCondition=GPT-4o-mini2026.0395.412.391.49595.695.694.2
Qwen3.5-9BCondition=Qwen3.5-9B2026.039523.784.696.592.896.377.5
Cascade (Q→14B)Condition=Cascade (Q→14B)2026.0394.79.792.596.592.895.6—
Qwen3.5-4BCondition=Qwen3.5-4B2026.0394.329.280.994.396.191.983.3
Cascade (Q→G)Condition=Cascade (Q→G)2026.0392.66.792.991.591.195.6—
BART-MNLICondition=BART-MNLI2026.038.10154.38.311.863.3
BaselineCondition=Baseline2026.030000000