Data as of Aug 16, 2026 · Based on 3,131,739 AI responses across 10,525 prompts · See how Parse measures this
Redwood Researchredwoodresearch.org · Low evidenceRedwood Research develops protocols to improve AI safety against intentional subversion. Their work includes testing strategies to detect hidden backdoors in code using models like GPT-4 as stand-ins for potentially deceptive AI.
Parse Score