Intent and Guide-Text Generation on GuideWeb
44.94Intent BLEUGuideWeb Agent
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GuideWeb Agent2026.02 | 44.94 | 52.89 | 21.34 | 28.44 | |
| Gemini-2.5-Pro2026.02 | 16.53 | 30.54 | 2.53 | 13.43 | |
| Claude-Sonnet-4.52026.02 | 15.88 | 25.99 | 1.48 | 8.98 | |
| GPT-52026.02 | 13.84 | 24.99 | 4.3 | 11.69 | |
| LLaMA3.1-8B2026.02 | 4.13 | 13.57 | 2.05 | 11.45 | |
| Qwen3-8B2026.02 | 3.69 | 11.54 | 1.93 | 11.47 |