| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Targeted Attack | Claude-Opus closed-source standard MLLMs 4.6 | ASR12.1 | 8 | |
| Text Detectability | claude-opus balanced 4-1 (test) | fastText F1_D98 | 5 | |
| Adversarial Attack | Claude-Opus 4.6 | ASR74 | 4 | |
| Adversarial Attack | Claude Opus 4.1 (test) | Bin-ASR46.7 | 3 |