How to Create a Weighted Index for AI Model Benchmark Results

Start with two complete sets of scores: one from your automated LLM-as-judge and one from your manual 'vibe check'. Ensure you have parallel scores for each model's output on every task.

My actual values:
[insert the files, settings, accounts, or constraints for this step]

Give me the exact commands, settings, or output to use. Finish with a pass or fail check for collate automated and manual scores.

Place the automated and manual leaderboards side by side. Identify which models ranked high on both, and which ones showed significant divergence. This helps reveal what the automated judge values versus your own taste.

My actual values:
[insert the files, settings, accounts, or constraints for this step]

Give me the exact commands, settings, or output to use. Finish with a pass or fail check for analyze score divergence.

I have two sets of benchmark scores for several AI models: one from an automated judge and one from my manual review. Generate the code for a single, self-contained HTML file that lets me create a weighted index from these two data sources. It should include a slider to adjust the weight from 0% to 100% between the manual and automated scores. The page should display a table with the final weighted rankings, which updates in real time as I move the slider. Here is the data: [paste your two sets of scores].

Move the slider on your tool to experiment with different weightings. Decide on the balance that best reflects your priorities. The blog author chose a 70% weight for their manual score and 30% for the automated score to create their final ranking.

My actual values:
[insert the files, settings, accounts, or constraints for this step]

Give me the exact commands, settings, or output to use. Finish with a pass or fail check for set your weights and determine final ranking.

Act as a data analyst. I'm providing you with the final weighted scores from my AI model benchmark. The data includes scores for each model across several distinct tasks: [list your tasks, e.g., writing PRDs, prototyping, agentic voice]. Based on this data, generate a summary of task-specific recommendations. For each task, identify the top-performing model and explain why it's the best choice for that specific job. Here is the data: [paste your final weighted score data].