Spatial Reasoning on GTB
32.69ScoreHSRL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| HSRL2026.05 | 32.69 | 44.41 | |
| Reflexion2026.05 | 29.34 | 37.76 | |
| System-1.x2026.05 | 27.73 | 30.01 | |
| CoT2026.05 | 26.58 | 31.61 | |
| HyperTree2026.05 | 25.81 | 26.67 | |
| Tree Planner2026.05 | 25.28 | 25.46 | |
| Direct Answer2026.05 | 23.61 | 25.96 | |
| ProgPrompt2026.05 | 22.45 | 26.12 | |
| ReAct2026.05 | 20.4 | 39.85 | |
| Inner Monologue2026.05 | 19.18 | 21.41 |