Classification on AI-READI Scenario B — Negative-Class Undersampling (train)
17.16AccuracyLLaMA*
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| LLaMA*Augmentation=CF Added, Prompting=Fine-tuned2026.01 | 17.16 | 16.13 | 16.39 | 16.39 | 23.53 | |
| BioMistral*Augmentation=CF Added, Prompting=Fine-tuned2026.01 | 14.82 | 14.52 | 13.11 | 14.75 | 20.59 | |
| GPT-4Augmentation=CF Added, Prompting=Zero-shot2026.01 | 6.86 | 4.84 | 4.92 | 4.92 | 8.82 | |
| LLaMAAugmentation=CF Added, Prompting=Zero-shot2026.01 | 5.93 | 3.23 | 3.28 | 3.28 | 5.88 | |
| BioMistralAugmentation=CF Added, Prompting=Zero-shot2026.01 | 4.52 | 1.61 | 1.64 | 1.64 | 4.41 | |
| NNData Condition=Model trained on reduced data2026.01 | -10.72 | -8.82 | -10.29 | -10.29 | -12.82 |