Deep research report generation on PDR-Bench
7.21P-ScoreIntentRL clarify
Evaluation results
| Method | Links | ||
|---|---|---|---|
| IntentRL clarifyAgent=Tavily Deep Research Agent2026.02 | 7.21 | 7.64 | |
| CollabLLM clarifyAgent=Tavily Deep Research Agent2026.02 | 7.12 | 7.6 | |
| IntentRL clarifyAgent=Gemini Deep Research Agent2026.02 | 7.06 | 7.48 | |
| Tell Me More clarifyAgent=Tavily Deep Research Agent2026.02 | 7.01 | 7.48 | |
| Learn-to-Ask clarifyAgent=Tavily Deep Research Agent2026.02 | 6.91 | 7.48 | |
| Tell Me More clarifyAgent=Gemini Deep Research Agent2026.02 | 6.84 | 7.24 | |
| CollabLLM clarifyAgent=Gemini Deep Research Agent2026.02 | 6.83 | 7.3 | |
| Learn-to-Ask clarifyAgent=Gemini Deep Research Agent2026.02 | 6.79 | 7.42 | |
| no clarifyAgent=Tavily Deep Research Agent2026.02 | 6.62 | 7.28 | |
| no clarifyAgent=Gemini Deep Research Agent2026.02 | 6.52 | 7.19 | |
| IntentRL clarifyAgent=Qwen Deep Research Agent2026.02 | 6.17 | 6.99 | |
| CollabLLM clarifyAgent=Qwen Deep Research Agent2026.02 | 6.11 | 6.92 | |
| IntentRL clarifyAgent=OpenAI Deep Research Agent2026.02 | 6.1 | 6.55 | |
| Learn-to-Ask clarifyAgent=Qwen Deep Research Agent2026.02 | 6.08 | 6.81 | |
| CollabLLM clarifyAgent=OpenAI Deep Research Agent2026.02 | 6.05 | 6.54 | |
| Tell Me More clarifyAgent=Qwen Deep Research Agent2026.02 | 6.02 | 6.73 | |
| Tell Me More clarifyAgent=OpenAI Deep Research Agent2026.02 | 6 | 6.39 | |
| OpenAI clarifyAgent=OpenAI Deep Research Agent2026.02 | 5.99 | 6.36 | |
| Learn-to-Ask clarifyAgent=OpenAI Deep Research Agent2026.02 | 5.85 | 6.51 | |
| no clarifyAgent=Qwen Deep Research Agent2026.02 | 5.73 | 6.71 | |
| no clarifyAgent=OpenAI Deep Research Agent2026.02 | 5.29 | 6.26 | |
| Qwen clarifyAgent=Qwen Deep Research Agent2026.02 | 4.59 | 5.29 |