List your company

Alignment defense against harmful fine-tuning on GSM8K

0.91Harmful Score (HS)Antibody
-0.58329.495919.57529.6541Feb 28, 2026

Evaluation results

MethodLinks
AntibodyBackbone=Gemma-2-9B2026.020.9157.43
LisaBackbone=Gemma-2-9B2026.0213.0254.97
BoosterBackbone=Gemma-2-9B2026.0222.1358.97
SFTBackbone=Gemma-2-9B2026.0232.0556.73
VaccineBackbone=Gemma-2-9B2026.0238.2454.37