Multi-Turn Retrieval-Augmented Generation on MTRAGEval Task B (test)
78.27HM3 (conditional)Our ensemble (1st place)
Evaluation Results
| Method | Links | |
|---|---|---|
| Our ensemble (1st place)Description=Ensemble of seven LLMs selected by GPT-4o-mini2026.05 | 78.27 | |
| gpt-oss-120bType=Baseline2026.05 | 63.9 |