Embodied Task Completion on ALFWorld (OOD)
97.25AccuracySKILLGEN
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SKILLGENModel Name=GPT-5.4-Mini, Skill Augmentation Protocol=SKILL2026.05 | 97.25 | 3.92 | |
| SKILLGENModel Name=GPT-5.4-Nano, Skill Augmentation Protocol=SKILL2026.05 | 94.9 | 21.96 | |
| SKILLGENModel Name=Gemma-4-26B, Skill Augmentation Protocol=SKILL2026.05 | 93.73 | 11.37 | |
| SKILLGENModel Name=Qwen-2.5-7B, Skill Augmentation Protocol=SKILL2026.05 | 82.35 | 12.94 | |
| SKILLGENModel Name=Grok-4-Fast, Skill Augmentation Protocol=SKILL2026.05 | 80.39 | 15.29 | |
| SKILLGENModel Name=Mistral-Nemo, Skill Augmentation Protocol=SKILL2026.05 | 67.06 | 8.24 | |
| SKILLGENModel Name=Llama-3.1-8B, Skill Augmentation Protocol=SKILL2026.05 | 65.1 | -2.35 | |
| SKILLGENModel Name=Claude-Haiku-4.5, Skill Augmentation Protocol=SKILL2026.05 | 64.71 | 3.53 |