Early mistake detection on Captain-Cook4D and Assembly101 (test)
56.8APMistExit
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MistExitAveraging=3 random seeds2026.03 | 56.8 | 3.7 | |
| FastForwardAveraging=3 random seeds2026.03 | 56.2 | 5 |
Loading the SOTA2 catalog…
| Method | Links | ||
|---|---|---|---|
| MistExitAveraging=3 random seeds2026.03 | 56.8 | 3.7 | |
| FastForwardAveraging=3 random seeds2026.03 | 56.2 | 5 |