Question Answering on MS MARCO
1Answerability RateVanilla RAG
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Vanilla RAG2026.05 | 1 | — | — | |
| RobustRAG2026.05 | 1 | — | — | |
| Paraphrase2026.05 | 1 | — | — | |
| Debate2026.05 | 1 | — | — | |
| TrustRAG2026.05 | 0.95 | — | — | |
| Cordon-MAS2026.05 | 0.79 | — | — | |
| Baseline2026.03 | — | 23.5 | — | |
| Cordon-MASLLM Backbone=DeepSeek2026.05 | — | — | 0.81 | |
| GMTP2026.03 | — | 24.5 | — | |
| GRADA2026.03 | — | 25.1 | — | |
| ProGRank2026.03 | — | 20.2 | — | |
| RAGuard2026.03 | — | 23.5 | — | |
| Vanilla RAGLLM Backbone=DeepSeek2026.05 | — | — | 0.67 |