Multi-step dependency reasoning on Minecraft
40WeaponSet Success@0→10Echo
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| EchoModel category=Few-shot Variants, k-shot=8, Component configuration=Full model2026.04 | 40 | 77.5 | 42.5 | 82.5 | 27.5 | 67.5 | |
| JARVIS-1Model category=Memory-enhanced Models, Component configuration=Full model2026.04 | 37.5 | 72.5 | 42.5 | 75 | 30 | 65 | |
| EchoModel category=Few-shot Variants, k-shot=1, Component configuration=Full model2026.04 | 37.5 | 72.5 | 32.5 | 67.5 | 20 | 60 | |
| EchoModel category=Few-shot Variants, k-shot=2, Component configuration=Full model2026.04 | 37.5 | 72.5 | 40 | 70 | 22.5 | 67.5 | |
| EchoModel category=Few-shot Variants, k-shot=4, Component configuration=Full model2026.04 | 37.5 | 72.5 | 40 | 75 | 22.5 | 67.5 | |
| MP5Model category=Memory-enhanced Models, Component configuration=Full model2026.04 | 35 | 65 | 35 | 62.5 | 30 | 57.5 | |
| PatrollerModel category=Memory-enhanced Models, Component configuration=Component disabled2026.04 | 27.5 | 50 | 25 | 45 | 22.5 | 42.5 | |
| SelfCheckModel category=Memory-enhanced Models, Component configuration=Component disabled2026.04 | 25 | 50 | 27.5 | 57.5 | 17.5 | 40 | |
| VoyagerModel category=Memory-enhanced Models, Component configuration=Full model2026.04 | 22.5 | 45 | 27.5 | 55 | 17.5 | 40 | |
| Self VerificationModel category=Memory-enhanced Models, Component configuration=Component disabled2026.04 | 15 | 32.5 | 17.5 | 40 | 10 | 27.5 | |
| MrSteveModel category=Memory-enhanced Models, Component configuration=Full model2026.04 | 15 | 32.5 | 17.5 | 35 | 12.5 | 27.5 |