Task Progress Estimation on Alfred
2.19pmaeR2VLM
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| R2VLMSize=7B2026.03 | 2.19 | 2.17 | 91.7 | 98.8 | |
| R2VLM-SFTSize=7B2026.03 | 2.77 | 2.27 | 87.6 | 96.1 | |
| GVL-SFTSize=7B2026.03 | 6.21 | 8.39 | 83 | 93 | |
| R2VLM-ZeroSize=7B2026.03 | 7.21 | 7.78 | 74.2 | 49 | |
| Gemini-2.5-Pro2026.03 | 16.27 | 17.37 | 48.1 | 83 | |
| GPT-52026.03 | 18.35 | 15.48 | 50.5 | 90.1 | |
| GLM-4.1V-ThinkingSize=9B2026.03 | 22.67 | 17.47 | 32.9 | 54.8 | |
| Qwen2.5-VLSize=72B2026.03 | 24.88 | 18.68 | 34.2 | 78.1 | |
| Qwen3-VL-Plus2026.03 | 27.3 | 15.29 | 37.9 | 77.7 | |
| MiniCPM-V-2.6Size=8B2026.03 | 27.47 | 22.04 | 47.8 | 37.3 | |
| Qwen2.5-VLSize=7B2026.03 | 27.87 | 10.67 | 29.5 | 49.4 | |
| ROVER2026.03 | 29.43 | 11.52 | 30.8 | 57.5 | |
| InternVL3Size=78B2026.03 | 30.54 | 16.65 | 36 | 76.5 | |
| LIVSize=<1B2026.03 | 35.87 | 32.44 | 15.3 | 49 | |
| InternVL3Size=8B2026.03 | 36.09 | 13.56 | 31.1 | 62.1 |