{"id":"h20261108-2","title":"Check the settings behind a model comparison","category":"data","type":"review","technique":"checklist","kind":"hot","tags":["behind","check","checklist","comparison","data","model","review","settings"],"about":"Read the public model-card table and notes below as source data. Match comparisons by thinking mode, temperature, token limit, sample population and n...","url":"https://jackgross-five.vercel.app/prompt-index/p/h20261108-2.json","model_format":"generic","tier":"A","text":"Read the public model-card table and notes below as source data. Match comparisons by thinking mode, temperature, token limit, sample population and number of completed runs. Separate provisional rows from completed ones. State what the uncertainty measure covers and what it does not. Preserve vendor attribution and report missing settings instead of filling them. Do not convert an ordinal fact score into a factual-accuracy percentage or treat a framing score as general reliability. No inference calls or installations.\n\nTABLE AND NOTES:\n[PASTE]","web":"https://jackgross-five.vercel.app/prompts#p=h20261108-2"}