Evidence-Anchored Consolidation on Evidence-Anchored Consolidation Evaluation
3.7Omission Error RateChatGPT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ChatGPTExecution Strategy=PECII2026.04 | 3.7 | 0.07 | 4.8 | |
| LLaMA-8BExecution Strategy=PECII2026.04 | 3.9 | 0.07 | 4.9 | |
| Qwen-4BExecution Strategy=PECII2026.04 | 4.3 | 0.08 | 4.8 | |
| LLaMA-3BExecution Strategy=PECII2026.04 | 4.8 | 0.08 | 4.8 | |
| Qwen-1.5BExecution Strategy=PECII2026.04 | 5.4 | 0.09 | 4.7 | |
| LLaMA-1BExecution Strategy=PECII2026.04 | 5.9 | 0.09 | 4.7 | |
| ChatGPTExecution Strategy=Standard Chunking2026.04 | 13.6 | 0.22 | 4.6 | |
| LLaMA-8BExecution Strategy=Standard Chunking2026.04 | 14.9 | 0.23 | 4.5 | |
| Qwen-4BExecution Strategy=Standard Chunking2026.04 | 16.8 | 0.25 | 4.3 | |
| LLaMA-3BExecution Strategy=Standard Chunking2026.04 | 19.2 | 0.27 | 4.1 | |
| Qwen-1.5BExecution Strategy=Standard Chunking2026.04 | 22.1 | 0.3 | 3.7 | |
| ChatGPTExecution Strategy=No Chunking (Baseline)2026.04 | 22.8 | 0.36 | 4.2 | |
| LLaMA-1BExecution Strategy=Standard Chunking2026.04 | 23.5 | 0.32 | 3.8 | |
| LLaMA-8BExecution Strategy=No Chunking (Baseline)2026.04 | 24.1 | 0.38 | 4 | |
| Qwen-4BExecution Strategy=No Chunking (Baseline)2026.04 | 27.4 | 0.41 | 3.7 | |
| LLaMA-3BExecution Strategy=No Chunking (Baseline)2026.04 | 30.6 | 0.44 | 3.4 | |
| Qwen-1.5BExecution Strategy=No Chunking (Baseline)2026.04 | 35.2 | 0.49 | 3 | |
| LLaMA-1BExecution Strategy=No Chunking (Baseline)2026.04 | 36.8 | 0.51 | 3.1 |