Belief assessment on misinformation on MISBELIEF Sports domain Medium
1.79Belief ScoreGPT-5
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5Setting=Baseline2026.01 | 1.79 | |
| GPT-5Setting=Original2026.01 | 1.91 | |
| Llama3-8BSetting=Baseline2026.01 | 1.92 | |
| Llama3-8BSetting=Original2026.01 | 1.96 | |
| Qwen2.5-32BSetting=Original2026.01 | 2.06 | |
| Qwen-turbo (w/o reasoning)Setting=Original2026.01 | 2.15 | |
| Qwen-turboSetting=Original2026.01 | 2.2 | |
| Qwen2.5-32BSetting=Baseline2026.01 | 2.31 | |
| Qwen-turbo (w/o reasoning)Setting=Baseline2026.01 | 2.4 | |
| Qwen-turboSetting=Baseline2026.01 | 2.44 | |
| GPT-3.5-turboSetting=Original2026.01 | 2.52 | |
| GPT-3.5-turboSetting=Baseline2026.01 | 2.64 | |
| Qwen2.5-72BSetting=Original2026.01 | 2.65 | |
| Qwen2.5-72BSetting=Baseline2026.01 | 2.88 | |
| Qwen-turbo (w/o reasoning)Setting=Round 12026.01 | 3.25 | |
| Llama3-8BSetting=Round 12026.01 | 3.38 | |
| Qwen-turbo (w/o reasoning)Setting=Round 32026.01 | 3.42 | |
| GPT-5Setting=Round 12026.01 | 3.53 | |
| Llama3-8BSetting=Round 32026.01 | 3.65 | |
| Qwen2.5-32BSetting=Round 12026.01 | 3.65 | |
| Qwen2.5-72BSetting=Round 12026.01 | 3.71 | |
| Qwen2.5-32BSetting=Round 32026.01 | 3.81 | |
| Qwen2.5-72BSetting=Round 32026.01 | 3.83 | |
| GPT-5Setting=Round 32026.01 | 3.87 | |
| Qwen-turboSetting=Round 12026.01 | 3.88 | |
| GPT-3.5-turboSetting=Round 12026.01 | 4.01 | |
| Qwen-turboSetting=Round 32026.01 | 4.05 | |
| GPT-3.5-turboSetting=Round 32026.01 | 4.16 |