Preference-aligned decision making on Mobile Manipulation (test)
67.1AccuracyAdaptive CLIPR
Evaluation Results
| Method | Links | |
|---|---|---|
| Adaptive CLIPRFeedback=Adaptive2026.05 | 67.1 | |
| CLIPRFeedback=Iterative learning approach2026.05 | 66.3 | |
| GATE (15-turn)Interaction turns=152026.05 | 58.8 | |
| CIPHER (Sem.)Distance Metric=Semantic2026.05 | 58.4 | |
| CIPHER (Lev.)Distance Metric=Levenshtein2026.05 | 53.7 | |
| Zero-shotProtocol=Zero-shot2026.05 | 33.8 | |
| ICLMode=In-context learning2026.05 | 33.3 | |
| TidyBot2026.05 | 31.6 | |
| IPMethod Identity=Introspective Planning2026.05 | 30.9 | |
| ICL + AnswersMode=In-context learning with answers2026.05 | 28.2 |