Interactive Agent Task Completion on ALFWorld unseen (val)
70.8Pick Success RateEvoSkill
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| EvoSkillBackbone=Qwen2.5-7B-Instruct2026.05 | 70.8 | 27.8 | 41.9 | 39.1 | 19 | 11.8 | 37.3 | |
| SkillBrewBackbone=Qwen2.5-7B-Instruct2026.05 | 70.8 | 72.2 | 58.1 | 60.9 | 57.1 | 29.4 | 59 | |
| Skill-ProBackbone=Qwen2.5-7B-Instruct2026.05 | 66.7 | 66.7 | 45.2 | 43.5 | 47.6 | 23.5 | 49.3 | |
| SimpleMemBackbone=Qwen2.5-7B-Instruct2026.05 | 64.5 | 33.3 | 20 | 12.5 | 33.3 | 3.84 | 29.7 | |
| VoyagerBackbone=Qwen2.5-7B-Instruct2026.05 | 62.5 | 61.1 | 51.6 | 52.2 | 28.6 | 17.6 | 47 | |
| ReflexionBackbone=Qwen2.5-7B-Instruct2026.05 | 62 | 41.6 | 44.9 | 30.9 | 36.3 | 23.8 | 42.7 | |
| MemPBackbone=Qwen2.5-7B-Instruct2026.05 | 54.3 | 38.5 | 48.1 | 56.2 | 32 | 16.7 | 41.4 | |
| Mem0Backbone=Qwen2.5-7B-Instruct2026.05 | 54 | 55 | 26.9 | 36.4 | 20.8 | 7.69 | 33.6 | |
| ReActBackbone=Qwen2.5-7B-Instruct2026.05 | 48.5 | 35.4 | 34.3 | 13.2 | 18.2 | 17.6 | 31.2 | |
| Zero-ShotBackbone=Qwen2.5-7B-Instruct2026.05 | 33.4 | 21.6 | 19.3 | 6.9 | 2.8 | 3.2 | 14.8 | |
| ExpeLBackbone=Qwen2.5-7B-Instruct2026.05 | 21 | 67 | 55 | 52 | 71 | 6 | 46.3 |