Long-horizon planning on ZebraLogic
93.5Accuracy
GRPO + coupled rhythm credit
Updated 1mo ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2025.10 | 93.5 | 9.6 | |
2025.10 | 93.2 | 9.3 | |
2025.10 | 91.4 | 7.5 | |
2025.10 | 90.3 | 6.4 | |
2025.10 | 89.3 | 5.4 | |
2025.10 | 83.9 | - |
GRPO + coupled rhythm credit
| Method | Links | ||
|---|---|---|---|
2025.10 | 93.5 | 9.6 | |
2025.10 | 93.2 | 9.3 | |
2025.10 | 91.4 | 7.5 | |
2025.10 | 90.3 | 6.4 | |
2025.10 | 89.3 | 5.4 | |
2025.10 | 83.9 | - |