Information Extraction on HFPP (test)
99.51State AccuracyFew-Shot
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Few-ShotEvaluation Protocol=Few-Shot, Backbone=Llama-3.3-70B-Instruct-Turbo2026.04 | 99.51 | 99.34 | 81.41 | 19.71 | 60.81 | 18.2 | 21.22 | |
| Ours (Llama-3.3)Backbone=Llama-3.3, Protocol=Role-based framework2026.04 | 99.51 | 99.49 | 89.47 | 44.16 | 96.45 | 43.1 | 45.22 | |
| Chain-of-ThoughtEvaluation Protocol=Chain-of-Thought, Backbone=Llama-3.3-70B-Instruct-Turbo2026.04 | 99.18 | 99.34 | 88.76 | 27.24 | 61.32 | 26.8 | 27.68 | |
| Zero-ShotEvaluation Protocol=Zero-Shot, Backbone=Llama-3.3-70B-Instruct-Turbo2026.04 | 99.01 | 98.19 | 88.16 | 22.58 | 61.07 | 21.05 | 24.11 | |
| Ours (Qwen-3)Backbone=Qwen-3, Protocol=Role-based framework2026.04 | 99.01 | 98.85 | 58.55 | 35.36 | 35.36 | 35.36 | 54.11 |