Alternative Annotator Test on Wiki
0.57Doc Fit (Adv Prob)PROXANN
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| PROXANNLLM Backbone=Llama-3.3-70B2025.07 | 0.57 | 0.67 | 0.58 | 0.5 | |
| PROXANNLLM Backbone=GPT-4o2025.07 | 0.56 | 0.68 | 0.66 | 0.55 | |
| PROXANNLLM Backbone=Qwen-3 8B2025.07 | 0.56 | 0.58 | 0.46 | 0.39 | |
| PROXANNLLM Backbone=Qwen-3 32B2025.07 | 0.55 | 0.63 | 0.47 | 0.42 | |
| PROXANNLLM Backbone=Qwen-2.5-72B2025.07 | 0.52 | 0.68 | 0.66 | 0.46 | |
| PROXANNLLM Backbone=Llama-3.1-8B2025.07 | 0.22 | 0.36 | 0.05 | 0.11 |