Loading blueprint versions...
Please wait while we gather all the unique runs for this blueprint.
Please wait while we gather all the unique runs for this blueprint.
Please wait while we prepare the detailed comparison.
DTEF: Predict response distributions for Islam. Source: Global Dialogues GD6 (https://github.com/collect-intel/global-dialogues)
Average key point coverage extent for each model across all prompts.
| Prompts vs. Models | Claude Haiku 4.5 | Gemini 2.5 Flash | Gemma 3 12b It | GPT 4.1 Mini | GPT 4.1 Nano | GPT 4o Mini | |
|---|---|---|---|---|---|---|---|
| Score | 3rd 85.8% | 4th 85.0% | 6th 78.5% | 2nd 87.1% | 5th 82.2% | 1st 87.7% | |
| 88.0% | 84% | 97% | 70% | 92% | 91% | 94% | |
| 88.5% | 88% | 89% | 81% | 92% | 91% | 90% | |
| 81.3% | 83% | 90% | 83% | 79% | 68% | 85% | |
| 78.7% | 74% | 81% | 83% | 82% | 70% | 82% | |
| 78.0% | 96% | 65% | 55% | 79% | 93% | 80% | |
| 87.2% | 90% | 88% | 86% | 89% | 79% | 91% | |
| 88.2% | 88% | 89% | 93% | 87% | 85% | 87% | |
| 80.8% | 79% | 81% | 76% | 86% | 76% | 87% | |
| 85.7% | 92% | 91% | 74% | 91% | 76% | 90% | |
| 87.5% | 84% | 79% | 84% | 94% | 93% | 91% |