Opinion Alignment on OQA
91.6Opinion AlignmentClaude-3.5-v2
Evaluation results
| Method | Links | |
|---|---|---|
| Claude-3.5-v2Prompt Type=Sociodemographic prompt, Elicitation Method=Verbalized, Calibration Status=Calibrated2025.07 | 91.6 | |
| Mistral-largePrompt Type=Base prompt, Elicitation Method=Paraphrase, Calibration Status=Calibrated2025.07 | 89.9 | |
| Qwen-2.5-72BPrompt Type=Base prompt, Elicitation Method=Paraphrase, Calibration Status=Calibrated2025.07 | 89.7 | |
| Llama-3.2-90BPrompt Type=Base prompt, Elicitation Method=Paraphrase, Calibration Status=Calibrated2025.07 | 89.6 | |
| OLMo-2-7B-IPrompt Type=Base prompt, Elicitation Method=Paraphrase, Calibration Status=Calibrated2025.07 | 89 | |
| Claude-3.5-v2Prompt Type=Base prompt, Elicitation Method=Paraphrase, Calibration Status=Calibrated2025.07 | 88.8 | |
| Llama-3.2-90BPrompt Type=Base prompt, Elicitation Method=Paraphrase, Calibration Status=Original2025.07 | 79.3 | |
| Claude-3.5-v2Prompt Type=Base prompt, Elicitation Method=Paraphrase, Calibration Status=Original2025.07 | 70.5 |