Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Counting Task on Spatial and Temporal Simulation Tasks Episode-level

4.203Subtask Accuracy

Cortex (full harness)

2.550442.979473.40853.83753Jul 6, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2026.07
4.2034.2628.464
2026.07
3.954.1428.091
2026.07
3.7413.9957.736
2026.07
3.5254.297.815
2026.07
3.324.0567.376
2026.07
3.063.9457.006
2026.07
2.6143.7756.388