Contributed by the Risk Threshold Forecasting community.

When will 75% accuracy be reached on LAB-Bench Cloning Scenarios by a Claude Sonnet 4 scale model by Anthropic?

Current estimate
Top Key Factors