Belief assessment on misinformation on MISBELIEF Sports domain Easy split
3.32Belief ScoreQwen2.5-72B
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen2.5-72BSetting=Round 32026.01 | 3.32 | |
| GPT-3.5-turboSetting=Round 32026.01 | 3.27 | |
| Qwen2.5-32BSetting=Round 32026.01 | 3.23 | |
| Qwen-turboSetting=Round 32026.01 | 3.19 | |
| Qwen2.5-32BSetting=Round 12026.01 | 3.06 | |
| Qwen2.5-72BSetting=Round 12026.01 | 2.98 | |
| GPT-3.5-turboSetting=Round 12026.01 | 2.94 | |
| Qwen-turboSetting=Round 12026.01 | 2.9 | |
| GPT-5Setting=Round 32026.01 | 2.78 | |
| Llama3-8BSetting=Round 32026.01 | 2.6 | |
| Qwen-turbo (w/o reasoning)Setting=Round 32026.01 | 2.46 | |
| GPT-5Setting=Round 12026.01 | 2.38 | |
| Qwen-turbo (w/o reasoning)Setting=Round 12026.01 | 2.29 | |
| Llama3-8BSetting=Round 12026.01 | 2.15 | |
| Qwen2.5-72BSetting=Baseline2026.01 | 2.07 | |
| GPT-3.5-turboSetting=Baseline2026.01 | 1.99 | |
| Qwen2.5-72BSetting=Original2026.01 | 1.78 | |
| Qwen-turboSetting=Baseline2026.01 | 1.71 | |
| Qwen-turbo (w/o reasoning)Setting=Baseline2026.01 | 1.71 | |
| GPT-3.5-turboSetting=Original2026.01 | 1.6 | |
| Qwen2.5-32BSetting=Baseline2026.01 | 1.59 | |
| Qwen-turbo (w/o reasoning)Setting=Original2026.01 | 1.48 | |
| GPT-5Setting=Baseline2026.01 | 1.39 | |
| Qwen-turboSetting=Original2026.01 | 1.37 | |
| Qwen2.5-32BSetting=Original2026.01 | 1.3 | |
| GPT-5Setting=Original2026.01 | 1.23 | |
| Llama3-8BSetting=Original2026.01 | 1.23 | |
| Llama3-8BSetting=Baseline2026.01 | 1.2 |