Use manual grid search over temperature+system to optimize a API design decisions prompt on o1-mini against F1 score without regressing safety.
Use manual grid search over temperature+system to optimize a log anomaly detection prompt on DeepSeek-V3 against F1 score without regressing safety.
Use manual grid search over temperature+system to optimize a product requirement drafting prompt on Claude 3.7 Sonnet against F1 score without regressing safety.
Use manual grid search over temperature+system to optimize a threat modeling prompt on o3 against factuality without regressing safety.