Visual Metaphor Transfer on Curated Visual Metaphor Dataset (test)
9.31MC ScoreVisual Metaphor Transfer framework
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Visual Metaphor Transfer frameworkEvaluation Model=Gemini-3-pro, Variant=Full Model2026.02 | 9.31 | 8.97 | 8.76 | 5.68 | |
| Visual Metaphor Transfer frameworkEvaluation Model=Gemini-3-pro, Variant=Ablation 3 (without Phase 4)2026.02 | 9.14 | 8.47 | 8.33 | 5.63 | |
| Visual Metaphor Transfer frameworkEvaluation Model=Gemini-3-pro, Variant=Ablation 2 (without CBT)2026.02 | 8.91 | 8.09 | 7.58 | 5.61 | |
| Visual Metaphor Transfer frameworkEvaluation Model=Gemini-3-pro, Variant=Ablation 1 (without CBT and Phases 1-2)2026.02 | 8.79 | 8.03 | 7.63 | 5.59 | |
| Banana-proEvaluation Model=Gemini-3-pro2026.02 | 8.75 | 7.68 | 7.33 | 5.57 | |
| Visual Metaphor Transfer frameworkEvaluation Model=Claude-4.5, Variant=Full Model2026.02 | 8.73 | 8.61 | 8.36 | 5.68 | |
| Visual Metaphor Transfer frameworkEvaluation Model=Claude-4.5, Variant=Ablation 3 (without Phase 4)2026.02 | 8.62 | 8.38 | 8.19 | 5.63 | |
| Visual Metaphor Transfer frameworkEvaluation Model=GPT-5.2, Variant=Full Model2026.02 | 8.62 | 8.51 | 8.58 | 5.68 | |
| Visual Metaphor Transfer frameworkEvaluation Model=Claude-4.5, Variant=Ablation 2 (without CBT)2026.02 | 8.56 | 7.89 | 7.97 | 5.61 | |
| Visual Metaphor Transfer frameworkEvaluation Model=Claude-4.5, Variant=Ablation 1 (without CBT and Phases 1-2)2026.02 | 8.44 | 7.85 | 7.92 | 5.59 | |
| Visual Metaphor Transfer frameworkEvaluation Model=GPT-5.2, Variant=Ablation 3 (without Phase 4)2026.02 | 8.44 | 8.33 | 8.29 | 5.63 | |
| Visual Metaphor Transfer frameworkEvaluation Model=GPT-5.2, Variant=Ablation 2 (without CBT)2026.02 | 8.33 | 8.01 | 7.71 | 5.61 | |
| Visual Metaphor Transfer frameworkEvaluation Model=GPT-5.2, Variant=Ablation 1 (without CBT and Phases 1-2)2026.02 | 8.13 | 7.96 | 7.69 | 5.59 | |
| GPT-ImageEvaluation Model=Gemini-3-pro2026.02 | 8.08 | 7.59 | 7.47 | 5.63 | |
| Banana-proEvaluation Model=Claude-4.52026.02 | 8.08 | 7.42 | 7.74 | 5.57 | |
| GPT-ImageEvaluation Model=Claude-4.52026.02 | 7.95 | 7.39 | 7.51 | 5.63 | |
| Banana-proEvaluation Model=GPT-5.22026.02 | 7.95 | 7.77 | 7.37 | 5.57 | |
| GPT-ImageEvaluation Model=GPT-5.22026.02 | 7.71 | 7.65 | 7.54 | 5.63 | |
| MidjourneyEvaluation Model=Claude-4.52026.02 | 6.51 | 5.94 | 6.06 | 5.22 | |
| MidjourneyEvaluation Model=GPT-5.22026.02 | 6.33 | 6.46 | 6.24 | 5.22 | |
| BAGELEvaluation Model=GPT-5.22026.02 | 6.21 | 5.83 | 6.07 | 4.77 | |
| BAGELEvaluation Model=Claude-4.52026.02 | 6.05 | 5.58 | 5.95 | 4.77 | |
| MidjourneyEvaluation Model=Gemini-3-pro2026.02 | 5.33 | 5.57 | 6.09 | 5.22 | |
| BAGELEvaluation Model=Gemini-3-pro2026.02 | 5.17 | 4.55 | 5.05 | 4.77 |