List your company

Benign Compliance on XSTest Vanilla

98.4LC WinrateQwen2.5-14B-IT
8890.793.496.1May 8, 2026

Evaluation results

MethodLinks
Qwen2.5-14B-ITBackbone=Qwen2.5-14B-IT2026.0598.4
SELF-REDTEAMBackbone=Qwen2.5-14B-IT2026.0596.8
ABSBackbone=Qwen2.5-14B-IT2026.0596.8
Qwen2.5-3B-ITBackbone=Qwen2.5-3B-IT2026.0596
ABSBackbone=Qwen2.5-3B-IT2026.0595.2
SELF-REDTEAMBackbone=Qwen2.5-3B-IT2026.0593.2
Qwen2.5-7B-ITBackbone=Qwen2.5-7B-IT2026.0591.2
SELF-REDTEAMBackbone=Qwen2.5-7B-IT2026.0588.8
ABSBackbone=Qwen2.5-7B-IT2026.0588.4