Spatial Reasoning on Blocksworld 5-7
30.5Completion RateHSRL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| HSRL2026.05 | 30.5 | 21 | |
| System-1.x2026.05 | 27 | 14.5 | |
| Reflexion2026.05 | 15 | 8.5 | |
| CoT2026.05 | 10.5 | 8 | |
| ProgPrompt2026.05 | 9.5 | 6 | |
| ReAct2026.05 | 8 | 3 | |
| HyperTree2026.05 | 8 | 3.5 | |
| Tree Planner2026.05 | 7 | 4 | |
| Direct Answer2026.05 | 6.5 | 6 | |
| Inner Monologue2026.05 | 4 | 0 |