List your company

Anti-Survival on Anti-Survival (test)

52.3Behavioral Alignment Score (Baseline)System Prompt
45.33247.14148.9550.759Aug 11, 2025

Evaluation results

MethodLinks
System PromptModel=LLaVA-1.52025.0852.360.849.8
Steering VectorModel=LLaVA-1.52025.0852.361.241
VISORModel=LLaVA-1.52025.0852.360.237.2
System PromptModel=Idefics22025.0845.664.841.6
Steering VectorModel=Idefics22025.0845.662.531.3
VISORModel=Idefics22025.0845.667.534.4