Spatial Information on Spatial and Temporal Simulation Tasks Episode-level
3.543Subtask AccuracyCortex (full harness)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Cortex (full harness)Level=Episode-level2026.07 | 3.543 | 4.044 | 7.587 | |
| Cortex (w/o harness)Level=Episode-level2026.07 | 3.033 | 3.771 | 6.805 | |
| Cortex (harness on skills)Level=Episode-level2026.07 | 3.017 | 3.958 | 6.985 | |
| GPT-5Level=Episode-level2026.07 | 2.92 | 4.4 | 7.321 | |
| GeminiLevel=Episode-level2026.07 | 2.896 | 4.033 | 6.929 | |
| Cortex (baseline)Level=Episode-level2026.07 | 2.549 | 3.969 | 6.518 | |
| Qwen3-VL-8B-InstructLevel=Episode-level2026.07 | 2.548 | 3.92 | 6.468 |