Histology Question Answering on SlideBench BCNB
56.89AccuracyOurs (SparseLearn)
Evaluation results
| Method | Links | |
|---|---|---|
| Ours (SparseLearn)FLOPs=1.73T, Token pruning ratio=~ 60×, Evaluation Protocol=zero-shot2026.06 | 56.89 | |
| SlideChatFLOPs=133.3T, Constraint=Upper Bound2026.06 | 54.14 | |
| GPT-4oToken pruning ratio=~ 60×, Evaluation Protocol=zero-shot2026.06 | 41.43 | |
| MedDrFLOPs=1.70T, Token pruning ratio=~ 60×, Evaluation Protocol=zero-shot2026.06 | 33.67 | |
| Quilt-LLaVAFLOPs=1.70T, Token pruning ratio=~ 60×, Evaluation Protocol=zero-shot2026.06 | 32.19 | |
| LLaVA-MedFLOPs=1.70T, Token pruning ratio=~ 60×, Evaluation Protocol=zero-shot2026.06 | 30.1 | |
| Random BaselineToken pruning ratio=~ 60×2026.06 | 24.4 |