Fire safety query answering on Geometry-related fire safety queries (Check-stage, Model)
76.2CorrectnessCamel Baseline
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Camel Baseline2026.06 | 76.2 | 1.7 | 22.1 | 77.1 | |
| Enhanced-tool agent2026.06 | 70.4 | 12.1 | 17.5 | 76.5 | |
| Basic-tool agent2026.06 | 5.4 | 5.8 | 88.8 | 8.3 |