Faithfulness Evaluation on ConFiQA
81.4In-AccOCC-RAG-1.7B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OCC-RAG-1.7Bthinking mode=false2026.05 | 81.4 | 5 | |
| OCC-RAG-0.6Bthinking mode=false2026.05 | 79.9 | 5.2 | |
| Qwen3-14Bthinking mode=true2026.05 | 78.3 | 7.4 | |
| Qwen3-8Bthinking mode=true2026.05 | 77.6 | 6.9 | |
| Qwen3-8Bthinking mode=false2026.05 | 75.9 | 9.2 | |
| Qwen3-32Bthinking mode=true2026.05 | 75.8 | 8.5 | |
| Qwen3-4Bthinking mode=true2026.05 | 74.1 | 7.5 | |
| gemma-3-27b-itthinking mode=false2026.05 | 73 | 8 | |
| gemma-3-12b-itthinking mode=false2026.05 | 72 | 7.6 | |
| Qwen3-32Bthinking mode=false2026.05 | 72 | 11.5 | |
| Qwen3-1.7Bthinking mode=true2026.05 | 70.4 | 8.3 | |
| Qwen3-14Bthinking mode=false2026.05 | 70.4 | 13.1 | |
| gemma-3-4b-itthinking mode=false2026.05 | 69.8 | 8.9 | |
| Qwen3-4Bthinking mode=false2026.05 | 69.7 | 10.3 | |
| Qwen3-1.7Bthinking mode=false2026.05 | 64.8 | 12.7 | |
| Qwen3-0.6Bthinking mode=true2026.05 | 64.5 | 8.2 | |
| gemma-3-1b-itthinking mode=false2026.05 | 62.1 | 7.7 | |
| SmolLM3-3Bthinking mode=true2026.05 | 60.5 | 13.3 | |
| Qwen3-0.6Bthinking mode=false2026.05 | 59.7 | 9 | |
| SmolLM3-3Bthinking mode=false2026.05 | 58.6 | 15.4 | |
| Pleias-RAG-1.2Bthinking mode=false2026.05 | 37.3 | 25.3 |