Language-driven scene representation on ALFRED Object Shift [OS]
83.92F1 ScoreFalcon-7B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Falcon-7BApproach=CLM2026.05 | 83.92 | 66.38 | 2.34 | |
| Mistral-7BApproach=CLM2026.05 | 82.72 | 63.46 | 2.75 | |
| LLaMA 3.1-8BApproach=CLM2026.05 | 80.44 | 58.86 | 4.27 | |
| Vicuna-7BApproach=IM2026.05 | 80.03 | 51.67 | 4.4 | |
| AlpacaApproach=IM2026.05 | 77.27 | 43.98 | 4.64 | |
| Flan-T5Approach=Seq2Seq2026.05 | 76.67 | 47.43 | 3.48 | |
| T5Approach=Seq2Seq2026.05 | 68.17 | 36.99 | 4.51 |