Clarification Generation on DeepResearch Bench offline (test)
2.43Quality ScoreIntentRL
Evaluation Results
| Method | Links | |
|---|---|---|
| IntentRL2026.02 | 2.43 | |
| Learn-to-Askbaseline_type=proactive LLM baseline2026.02 | 2.28 | |
| CollabLLMbaseline_type=proactive LLM baseline2026.02 | 2.15 | |
| Tell Me Morebaseline_type=proactive LLM baseline2026.02 | 1.44 |