Continuous (Maze) Task on AMAZE
5.41ViolationJanus-Pro
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Janus-ProCoT=false, Fine-tuned=false2026.04 | 5.41 | 1.85 | 57.47 | 76.8 | 0 | 0 | |
| Janus-ProCoT=true, Fine-tuned=false2026.04 | 6.03 | 0.89 | 53.02 | 73.98 | 0 | 0 | |
| BagelCoT=false, Fine-tuned=true2026.04 | 12.21 | 51.02 | 8.66 | 3.07 | 11.54 | 23.64 | |
| BagelCoT=true, Fine-tuned=true2026.04 | 15.24 | 44.65 | 10.17 | 5.25 | 17.9 | 18.42 | |
| Seedream-4.5CoT=false2026.04 | 16.9 | 25.67 | 28.82 | 30.96 | 2.14 | 3.21 | |
| Qwen-Image-EditCoT=false2026.04 | 19.37 | 28.51 | 18.82 | 5.7 | 1.43 | 2.14 | |
| Flux-Kontext-DevCoT=false2026.04 | 23.84 | 30.24 | 30.96 | 18.31 | 0.36 | 3.57 | |
| BagelCoT=false, Fine-tuned=false2026.04 | 28.91 | 27.15 | 11.64 | 5.84 | 0 | 1 | |
| Janus-ProCoT=true, Fine-tuned=true2026.04 | 31.23 | 25.12 | 56.81 | 52.28 | 2.79 | 4.13 | |
| BagelCoT=true, Fine-tuned=false2026.04 | 34.06 | 30.31 | 14.77 | 3.97 | 0 | 0.57 | |
| Janus-ProCoT=false, Fine-tuned=true2026.04 | 35.6 | 23.33 | 55.99 | 50.94 | 1.43 | 2.22 | |
| NanoBanana-ProCoT=false2026.04 | 47.76 | 64.21 | 24.2 | 17.21 | 4.82 | 9.28 | |
| GPT-image-1CoT=false2026.04 | 62.88 | 58.97 | 41.16 | 52.76 | 5.4 | 6.06 |