# Classification on ColonINST (unseen)

Source: [Classification on ColonINST (unseen)](<https://www.sota2.com/research/sota/classification-on-coloninst-unseen>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:27:05.329Z

Evaluation of classification performance on the unseen split of the ColonINST dataset, reporting Accuracy.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | CLS |
| Dataset (including any stated split) | ColonINST (unseen) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 17 of 17 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Accuracy](<https://www.sota2.com/research/sota/classification-on-coloninst-unseen.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| ColonGPT | 83.24 | Not recorded | 68fa7b831e209131cea88cd0 |
| LLaVA-v1.5 | 80.89 | Not recorded | 68fa7b831e209131cea88cd0 |
| MobileVLM-1.7B | 80.44 | Not recorded | 68fa7b831e209131cea88cd0 |
| Bunny-v1.0-3B | 79.5 | Not recorded | 68fa7b831e209131cea88cd0 |
| LLaVA-Med-v1.5 | 79.24 | Not recorded | 68fa7b831e209131cea88cd0 |
| LLaVA-v1.5 | 79.1 | Not recorded | 68fa7b831e209131cea88cd0 |
| MGM-2B | 78.99 | Not recorded | 68fa7b831e209131cea88cd0 |
| MobileVLM-1.7B | 78.75 | Not recorded | 68fa7b831e209131cea88cd0 |
| MGM-2B | 78.69 | Not recorded | 68fa7b831e209131cea88cd0 |
| LLaVA-Med-v1.0 | 78.04 | Not recorded | 68fa7b831e209131cea88cd0 |
| MiniGPT-v2 | 77.93 | Not recorded | 68fa7b831e209131cea88cd0 |
| LLaVA-Med-v1.0 | 77.38 | Not recorded | 68fa7b831e209131cea88cd0 |
| MiniGPT-v2 | 76.82 | Not recorded | 68fa7b831e209131cea88cd0 |
| Bunny-v1.0-3B | 75.5 | Not recorded | 68fa7b831e209131cea88cd0 |
| LLaVA-v1 | 72.08 | Not recorded | 68fa7b831e209131cea88cd0 |
| LLaVA-Med-v1.5 | 66.51 | Not recorded | 68fa7b831e209131cea88cd0 |
| LLaVA-v1 | 42.17 | Not recorded | 68fa7b831e209131cea88cd0 |

### ColonGPT

Accuracy: 83.24. Result ID: 698b865d7a967eddb76f4f5d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Implementation](<https://github.com/ai4colonoscopy/IntelliScope>)

Recorded evaluation configuration: {"Visual encoder":"SigLIP-SO (384px)","Language model":"Phi1.5 (1.3B)","LoRA":false,"EXT":false}. This record does not assert independent reproduction.

### LLaVA-v1.5

Accuracy: 80.89. Result ID: 698b865d7a967eddb76f4f52.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa732e1e209131cea5af59.md>)

[Implementation](<https://llava-vl.github.io/>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (336px)","Language model":"Vicuna-v1.5 (7B)","LoRA":true,"EXT":true}. This record does not assert independent reproduction.

### MobileVLM-1.7B

Accuracy: 80.44. Result ID: 698b865d7a967eddb76f4f58.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71a71e209131cea52555.md>)

[Implementation](<https://github.com/Meituan-AutoML/MobileVLM>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (336px)","Language model":"MobileLLaMA (1.4B)","LoRA":true,"EXT":true}. This record does not assert independent reproduction.

### Bunny-v1.0-3B

Accuracy: 79.5. Result ID: 698b865d7a967eddb76f4f54.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74351e209131cea60750.md>)

[Implementation](<https://github.com/BAAI-DCAI/Bunny>)

Recorded evaluation configuration: {"Visual encoder":"SigLIP-SO (384px)","Language model":"Phi2 (2.7B)","LoRA":true,"EXT":true}. This record does not assert independent reproduction.

### LLaVA-Med-v1.5

Accuracy: 79.24. Result ID: 698b865d7a967eddb76f4f5b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8e1e209131cea46356.md>)

[Implementation](<https://aka.ms/llava-med>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (224px)","Language model":"Mistral-v0.2 (7B)","LoRA":true,"EXT":false}. This record does not assert independent reproduction.

### LLaVA-v1.5

Accuracy: 79.1. Result ID: 698b865d7a967eddb76f4f51.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa732e1e209131cea5af59.md>)

[Implementation](<https://llava-vl.github.io/>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (336px)","Language model":"Vicuna-v1.5 (7B)","LoRA":true,"EXT":false}. This record does not assert independent reproduction.

### MGM-2B

Accuracy: 78.99. Result ID: 698b865d7a967eddb76f4f55.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ab1e209131cea57bca.md>)

[Implementation](<https://github.com/dvlab-research/MiniGemini>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (336px) & ConvNeXt-L (768px)","Language model":"Gemma (2B)","LoRA":false,"EXT":false}. This record does not assert independent reproduction.

### MobileVLM-1.7B

Accuracy: 78.75. Result ID: 698b865d7a967eddb76f4f57.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71a71e209131cea52555.md>)

[Implementation](<https://github.com/Meituan-AutoML/MobileVLM>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (336px)","Language model":"MobileLLaMA (1.4B)","LoRA":true,"EXT":false}. This record does not assert independent reproduction.

### MGM-2B

Accuracy: 78.69. Result ID: 698b865d7a967eddb76f4f56.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ab1e209131cea57bca.md>)

[Implementation](<https://github.com/dvlab-research/MiniGemini>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (336px) & ConvNeXt-L (768px)","Language model":"Gemma (2B)","LoRA":true,"EXT":false}. This record does not assert independent reproduction.

### LLaVA-Med-v1.0

Accuracy: 78.04. Result ID: 698b865d7a967eddb76f4f59.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8e1e209131cea46356.md>)

[Implementation](<https://aka.ms/llava-med>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (224px)","Language model":"LLaMA1 (7B)","LoRA":false,"EXT":false}. This record does not assert independent reproduction.

### MiniGPT-v2

Accuracy: 77.93. Result ID: 698b865d7a967eddb76f4f4d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71251e209131cea4f7e5.md>)

[Implementation](<https://minigpt-v2.github.io/>)

Recorded evaluation configuration: {"Visual encoder":"EVA-G/14 (448px)","Language model":"LLaMA2 (7B)","LoRA":false,"EXT":false}. This record does not assert independent reproduction.

### LLaVA-Med-v1.0

Accuracy: 77.38. Result ID: 698b865d7a967eddb76f4f5a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8e1e209131cea46356.md>)

[Implementation](<https://aka.ms/llava-med>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (224px)","Language model":"LLaMA1 (7B)","LoRA":true,"EXT":false}. This record does not assert independent reproduction.

### MiniGPT-v2

Accuracy: 76.82. Result ID: 698b865d7a967eddb76f4f4e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71251e209131cea4f7e5.md>)

[Implementation](<https://minigpt-v2.github.io/>)

Recorded evaluation configuration: {"Visual encoder":"EVA-G/14 (448px)","Language model":"LLaMA2 (7B)","LoRA":true,"EXT":true}. This record does not assert independent reproduction.

### Bunny-v1.0-3B

Accuracy: 75.5. Result ID: 698b865d7a967eddb76f4f53.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74351e209131cea60750.md>)

[Implementation](<https://github.com/BAAI-DCAI/Bunny>)

Recorded evaluation configuration: {"Visual encoder":"SigLIP-SO (384px)","Language model":"Phi2 (2.7B)","LoRA":true,"EXT":false}. This record does not assert independent reproduction.

### LLaVA-v1

Accuracy: 72.08. Result ID: 698b865d7a967eddb76f4f4f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa717d1e209131cea51970.md>)

[Implementation](<https://llava-vl.github.io/>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (224px)","Language model":"Vicuna-v1.3 (7B)","LoRA":true,"EXT":false}. This record does not assert independent reproduction.

### LLaVA-Med-v1.5

Accuracy: 66.51. Result ID: 698b865d7a967eddb76f4f5c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8e1e209131cea46356.md>)

[Implementation](<https://aka.ms/llava-med>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (224px)","Language model":"Mistral-v0.2 (7B)","LoRA":false,"EXT":false}. This record does not assert independent reproduction.

### LLaVA-v1

Accuracy: 42.17. Result ID: 698b865d7a967eddb76f4f50.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7b831e209131cea88cd0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa717d1e209131cea51970.md>)

[Implementation](<https://llava-vl.github.io/>)

Recorded evaluation configuration: {"Visual encoder":"CLIP-L/14 (224px)","Language model":"Vicuna-v1.3 (7B)","LoRA":true,"EXT":true}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.