Context Extraction on SQuAD v2
89.3AccuracyBERT-Judge
Evaluation results
| Method | Links | |
|---|---|---|
| BERT-Judgeclean_name=BERT-as-a-Judge2026.04 | 89.3 | |
| Regex2026.04 | 72.3 | |
| LLM-Judgebackbone=Qwen-3 0.6B2026.04 | 62.5 |
| Method | Links | |
|---|---|---|
| BERT-Judgeclean_name=BERT-as-a-Judge2026.04 | 89.3 | |
| Regex2026.04 | 72.3 | |
| LLM-Judgebackbone=Qwen-3 0.6B2026.04 | 62.5 |