| image-classification-on-coloninst-v1-seen | LLaVA-Med-v1.0 (w/o LoRA, w/ extra data) | #2 | Accuray: 93.84 |
| image-classification-on-coloninst-v1-seen | LLaVA-Med-v1.5 (w/ LoRA, w/o extra data) | #4 | Accuray: 93.62 |
| image-classification-on-coloninst-v1-seen | LLaVA-Med-v1.0 (w/o LoRA, w/o extra data) | #5 | Accuray: 93.52 |
| image-classification-on-coloninst-v1-seen | LLaVA-Med-v1.5 (w/ LoRA, w/ extra data) | #17 | Accuray: 87.22 |
| image-classification-on-coloninst-v1-unseen | LLaVA-Med-v1.5
(w/ LoRA, w/o extra data) | #5 | Accuray: 79.24 |
| image-classification-on-coloninst-v1-unseen | LLaVA-Med-v1.0
(w/o LoRA, w/o extra data) | #10 | Accuray: 78.04 |
| image-classification-on-coloninst-v1-unseen | LLaVA-Med-v1.0
(w/o LoRA, w/ extra data) | #12 | Accuray: 77.38 |
| image-classification-on-coloninst-v1-unseen | LLaVA-Med-v1.5
(w/ LoRA, w/ extra data) | #16 | Accuray: 66.51 |
| referring-expression-generation-on-coloninst | LLaVA-Med-v1.5
(w/ LoRA, w/o extra data) | #3 | Accuray: 99.3 |
| referring-expression-generation-on-coloninst | LLaVA-Med-v1.0
(w/o LoRA, w/o extra data) | #9 | Accuray: 97.74 |
| referring-expression-generation-on-coloninst | LLaVA-Med-v1.0
(w/o LoRA, w/ extra data) | #10 | Accuray: 97.35 |
| referring-expression-generation-on-coloninst | LLaVA-Med-v1.5
(w/ LoRA, w/ extra data) | #14 | Accuray: 90.4 |
| referring-expression-generation-on-coloninst-1 | LLaVA-Med-v1.0
(w/o LoRA, w/ extra data) | #3 | Accuray: 75.25 |
| referring-expression-generation-on-coloninst-1 | LLaVA-Med-v1.0
(w/o LoRA, w/o extra data) | #5 | Accuray: 75.07 |
| referring-expression-generation-on-coloninst-1 | LLaVA-Med-v1.5
(w/ LoRA, w/o extra data) | #8 | Accuray: 73.05 |
| referring-expression-generation-on-coloninst-1 | LLaVA-Med-v1.5
(w/ LoRA, w/ extra data) | #13 | Accuray: 70.00 |