List your company

Safety on StrongReject

99.7Safety RateVanilla SFT
35.53252.19168.8585.509Feb 2, 2026

Evaluation results

MethodLinks
Vanilla SFTAlignment Dataset=DirectRefusal, Backbone=s1.1-7B2026.0299.7
Vanilla SFTAlignment Dataset=STAR-1, Backbone=s1.1-7B2026.0299.7
DGR SFTAlignment Dataset=DirectRefusal, Backbone=s1.1-7B2026.0298.7
Vanilla SFTAlignment Dataset=R1-ACT, Backbone=s1.1-7B2026.0298.1
DGR SFTAlignment Dataset=STAR-1, Backbone=s1.1-7B2026.0298.1
DGR SFTAlignment Dataset=R1-ACT, Backbone=s1.1-7B2026.0298.1
Qwen2.5-7B-Instruct2026.0294.6
s1.1-7B2026.0238