Question Answering on Average across QA benchmarks (test)
38.1EMR2-Searcher
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| R2-Searcher2026.06 | 38.1 | 45.8 | |
| AutoRefine2026.06 | 35.2 | 42.7 | |
| ReSearch2026.06 | 31.7 | 39.3 | |
| Search-R12026.06 | 31 | 38.3 | |
| ZeroSearch2026.06 | 27.3 | 34.3 | |
| Search-o12026.06 | 19.1 | 23.6 | |
| RAG2026.06 | 18.1 | 22.3 | |
| IRCOT2026.06 | 17.3 | 25.5 | |
| COT2026.06 | 14 | 18.7 | |
| Direct Inference2026.06 | 13.5 | 18 |