Guide target element selection on GuideWeb
29.99PrecisionGuideWeb Agent
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GuideWeb AgentMatch/Pred.=206/6872026.02 | 29.99 | 31.64 | 30.79 | |
| Claude-Sonnet-4.5Match/Pred.=384/24022026.02 | 15.99 | 58.99 | 25.16 | |
| GPT-5Match/Pred.=356/23282026.02 | 15.29 | 54.69 | 23.9 | |
| Gemini-2.5-ProMatch/Pred.=357/24742026.02 | 14.43 | 54.84 | 22.85 | |
| LLaMA3.1-8BMatch/Pred.=18/1642026.02 | 10.98 | 2.77 | 4.42 | |
| Qwen3-8BMatch/Pred.=72/7902026.02 | 9.11 | 12.33 | 10.48 |