Language-driven scene representation on ALFRED Template Shift [TS]
84.9F1 ScoreFalcon-7B
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Falcon-7BApproach=CLM2026.05 | 84.9 | 65.07 | 2.52 | |
| Mistral-7BApproach=CLM2026.05 | 83.97 | 62.93 | 2.75 | |
| LLaMA 3.1-8BApproach=CLM2026.05 | 81.87 | 59.65 | 4.08 | |
| Vicuna-7BApproach=IM2026.05 | 81.37 | 53.39 | 4 | |
| AlpacaApproach=IM2026.05 | 80.01 | 46.92 | 4.29 | |
| Flan-T5Approach=Seq2Seq2026.05 | 79.57 | 46.32 | 3.54 | |
| T5Approach=Seq2Seq2026.05 | 73.28 | 35.16 | 4.45 |