Sarcasm Understanding on MuSTARD
86.8PrecisionMuVaC
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MuVaC2026.01 | 86.8 | 89.2 | 88 | |
| Qwen 2.5 VLFine-tuned=true2026.01 | 67.5 | 70.3 | 68.9 | |
| Qwen 2.5 VLFine-tuned=false2026.01 | 61.9 | 41.4 | 49.6 | |
| Kimi VLFine-tuned=false2026.01 | 61.5 | 16.2 | 25.7 | |
| MiniCPM-V 2.6Fine-tuned=false2026.01 | 57.2 | 62.6 | 59.8 | |
| Qwen 2.5Fine-tuned=false2026.01 | 55.9 | 79.1 | 65.5 | |
| Llama 3.1Fine-tuned=false2026.01 | 53.8 | 83.1 | 65.3 |