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 GD7 (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 | 4th 80.8% | 3rd 81.0% | 6th 70.7% | 1st 83.3% | 2nd 81.3% | 5th 77.0% | |
| 84.7% | 93% | 87% | 80% | 83% | 79% | 86% | |
| 73.8% | 83% | 86% | 55% | 79% | 79% | 61% | |
| 70.3% | 73% | 69% | 61% | 73% | 73% | 73% | |
| 86.3% | 88% | 90% | 80% | 90% | 85% | 85% | |
| 80.5% | 88% | 85% | 68% | 76% | 88% | 78% | |
| 77.0% | 85% | 85% | 68% | 76% | 76% | 72% | |
| 85.7% | 87% | 87% | 73% | 87% | 92% | 88% | |
| 78.0% | 81% | 81% | 72% | 87% | 72% | 75% | |
| 79.0% | 69% | 73% | 75% | 92% | 87% | 78% | |
| 74.8% | 61% | 67% | 75% | 90% | 82% | 74% |