Actionability Filtering on CLIP
86.6Binary F1 ScoreTTP+Cxt
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| TTP+CxtModel Category=BERT Models2026.05 | 86.6 | — | — | — | — | 66.8 | |
| DNote+CxtModel Category=BERT Models2026.05 | 85.6 | — | — | — | — | 66.1 | |
| GPT-5.2Model Category=LLM Models, Prompting Strategy=Staged prompting strategy2026.05 | 81.3 | 84.4 | 78.5 | 90.9 | 88.3 | 89.5 | |
| Gemini-3-FlashModel Category=LLM Models, Prompting Strategy=Staged prompting strategy2026.05 | 80.1 | 80.3 | 79.8 | 88.9 | 88.7 | 88.8 | |
| Claude 3.5 SonnetModel Category=LLM Models, Prompting Strategy=Staged prompting strategy2026.05 | 73.5 | 73.1 | 73.9 | 84.9 | 85.3 | 85.1 | |
| DeepSeek-V3.2Model Category=LLM Models, Prompting Strategy=Staged prompting strategy2026.05 | 66.1 | 65.8 | 66.5 | 80.7 | 81.1 | 80.9 | |
| MedGemma-27B-itModel Category=LLM Models, Prompting Strategy=Staged prompting strategy2026.05 | 45.9 | 37.9 | 58.2 | 64.2 | 74.8 | 68.4 |