Visual Question Answering on Illusion (test)
98.41Overall AccuracyIllusion-aware VLM pipeline (Claude-opus-4-6)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Illusion-aware VLM pipeline (Claude-opus-4-6)Model=Claude-opus-4-6, Ensemble=5-vote majority, Pipeline=Full preprocessing2026.05 | 98.41 | — | — | |
| Illusion-aware VLM pipeline (Claude-opus-4-6)Model=Claude-opus-4-6, Ensemble=5-vote majority, Pipeline=Full preprocessing2026.05 | 90.48 | 82.38 | 98.57 |