List your company

Risk Detection on EMBHAZARD (Held-out Set)

85Potential Risk ScoreHuman
60.0466.527379.48May 29, 2026

Evaluation results

MethodLinks
Human2026.058592.858.8
EMBGUARD-4BParameters=4B2026.057760.753.3
Qwen-3-VL-235B-a22bParameters=235B2026.0570.371.746.2
GPT-5.12026.0567.367.261.9
EMBGUARD-2BParameters=2B2026.056139.222.7