Toxicity Detection (Stage I) on TOXICTAGS (test)
72.55mF1STEMTOX
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| STEMTOXmodel=PALIGEMMA, task/shot=multi2025.08 | 72.55 | 83.9 | |
| LHLmodel=PALIGEMMA, task/shot=multi2025.08 | 70.48 | 82.2 | |
| STEMTOXmodel=LLAVA, task/shot=multi2025.08 | 67.24 | 77.7 | |
| LHLmodel=LLAVA, task/shot=multi2025.08 | 65.03 | 76 | |
| FS (TS)model=GPT, task/shot=4-shot2025.08 | 63.37 | 78.87 | |
| STEMTOXmodel=LLAVA, task/shot=uni2025.08 | 62.19 | 77.2 | |
| LHLmodel=LLAVA, task/shot=uni2025.08 | 61.58 | 74.6 | |
| LHLmodel=PALIGEMMA, task/shot=uni2025.08 | 61.53 | 70.4 | |
| FS (TS)model=GPT, task/shot=2-shot2025.08 | 61.26 | 75.37 | |
| STEMTOXmodel=PALIGEMMA, task/shot=uni2025.08 | 60.85 | 74.9 |