Multi-step navigation and information location on BrowseComp English
54.9Score
GPT-5-High
Updated 1mo ago
Evaluation Results
| Method | Links | |
|---|---|---|
2026.03 | 54.9 | |
2026.03 | 52 | |
2026.03 | 51.5 | |
2026.03 | 51.4 | |
2026.03 | 49.1 | |
2026.03 | 45.1 | |
2026.03 | 44 | |
2026.03 | 43.4 | |
2026.03 | 35.3 | |
2026.03 | 30 | |
2026.03 | 29.5 | |
2026.03 | 27.7 | |
2026.03 | 24.4 | |
2026.03 | 24.1 | |
2026.03 | 18.8 | |
2026.03 | 15.3 | |
2026.03 | 14.7 | |
2026.03 | 14.1 | |
2026.03 | 13 | |
2026.03 | 10.6 | |
2026.03 | 10.5 | |
2026.03 | 3.8 |