Project {{model}} performance at {{target_scale}} based on current data. Step 1: Collect performance data at multiple scales from {{scaling_data}} Step 2: Fit scaling law curves to observed data Step 3: Extrapolate to target scale with confidence bounds Step 4: Identify which capabilities scale predictably Step 5: Flag emergent capability thresholds Step 6: Recommend scale investments based on projections Show mathematical reasoning at each step.
49 copies0 forks
Details
Category
AnalysisUse Cases
Scaling projectionPerformance forecastingInvestment planning
Works Best With
claude-opus-4.5gpt-5.2gemini-2.0-flash
Created Shared