List your company

Complex Reasoning on BBH (Solution Rate %)

67.4BBH Solution RateIT
63.65664.62865.666.572Feb 9, 2026

Evaluation results

MethodLinks
ITBase Model=Gemma2 9B2026.0267.4
RLBFBase Model=Gemma2 9B2026.0267.3
BSAFE+Base Model=Gemma2 9B2026.0267.1
RLBFBase Model=LLaMA 3 8B2026.0264.2
ITBase Model=LLaMA 3 8B2026.0264.1
BSAFE+Base Model=LLaMA 3 8B2026.0263.8