Classification on MAMI
89.1AccuracyGPT-4o
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-4oShots=8, Task Mode=Joint (C+E+I)2026.01 | 89.1 | 89.07 | |
| GPT-4oShots=4, Task Mode=Joint (C+E+I)2026.01 | 88.78 | 88.75 | |
| GPT-4oShots=2, Task Mode=Joint (C+E+I)2026.01 | 88.25 | 88.23 | |
| Vlm-Lim (GPT-4o)Task Mode=Single-task2026.01 | 83.6 | 83.54 | |
| LoReHM (GPT-4o)Task Mode=Single-task2026.01 | 83 | 82.98 | |
| M2KETask Mode=Single-task2026.01 | 75.85 | 75.71 | |
| PixtralShots=8, Task Mode=Joint (C+E+I)2026.01 | 75.42 | 74.99 | |
| Intern-VL3Shots=8, Task Mode=Joint (C+E+I)2026.01 | 75.34 | 75.25 | |
| Intern-VL3Shots=4, Task Mode=Joint (C+E+I)2026.01 | 74.39 | 74.24 | |
| PixtralShots=4, Task Mode=Joint (C+E+I)2026.01 | 73.65 | 72.89 | |
| Pro-CapTask Mode=Single-task2026.01 | 73.63 | 73.42 | |
| Intern-VL3Shots=2, Task Mode=Joint (C+E+I)2026.01 | 71.96 | 71.69 | |
| Few-ShotTask Mode=Single-task2026.01 | 70.5 | 70.1 | |
| PromptHateTask Mode=Single-task2026.01 | 70.31 | 70.18 | |
| Mod-HateTask Mode=Single-task2026.01 | 69.05 | 68.78 | |
| PixtralShots=2, Task Mode=Joint (C+E+I)2026.01 | 67.94 | 65.88 |