Papers with multi-turn retrieval augment generation
MTRAG-UN: A Benchmark for Open Challenges in Multi-Turn RAG Conversations (2026.findings-acl)
Copied to clipboard
| Challenge: | Several benchmarks have been released to evaluate model performance on multi-turn retrieval augment generation tasks. |
| Approach: | They propose to benchmark 666 conversations with over 2,800 conversation turns across 6 domains and a corpora that focuses on unanswerable questions and later conversation turns. |
| Outcome: | The proposed benchmarks show that retrieval and generation models struggle on conversations with UNanswerable, UNderspecified, and NONstandalone questions and UNclear responses. |