# 3D Text-Fidelity Preference Evaluation on OOD Preference Evaluation Dataset Text-Fidelity

Source: [3D Text-Fidelity Preference Evaluation on OOD Preference Evaluation Dataset Text-Fidelity](<https://www.sota2.com/research/sota/3d-text-fidelity-preference-evaluation-on-ood-preference-evaluation-dataset-text-fidelity>)

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

Data updated: 2026-02-26T05:30:29.363Z

Evaluation of 3D models based on human preference judgments concerning text fidelity, measured using Accuracy on the OOD Preference Evaluation Dataset Text-Fidelity.

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 | 3D Text-Fidelity Preference Evaluation |
| Dataset (including any stated split) | OOD Preference Evaluation Dataset Text-Fidelity |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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/3d-text-fidelity-preference-evaluation-on-ood-preference-evaluation-dataset-text-fidelity.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Gen3DEval (CLIP) | 86 | Not recorded | 68fa77f61e209131cea763ae |
| Image Reward Score | 85 | Not recorded | 68fa77f61e209131cea763ae |
| PickScore | 85 | Not recorded | 68fa77f61e209131cea763ae |
| CLIP Score | 80 | Not recorded | 68fa77f61e209131cea763ae |
| Gen3DEval (CLIP + DinoV2) | 74 | Not recorded | 68fa77f61e209131cea763ae |
| Phi-3.5-Vision | 65 | Not recorded | 68fa77f61e209131cea763ae |
| LLaVA-Qwen-7B | 58 | Not recorded | 68fa77f61e209131cea763ae |
| Gen3DEval w/ DinoV2 | 58 | Not recorded | 68fa77f61e209131cea763ae |
| GPT-4o | 55 | Not recorded | 68fa77f61e209131cea763ae |
| Gen3DEval (CLIP + Fit3D) | 53 | Not recorded | 68fa77f61e209131cea763ae |
| LLaVA-Llama3-8b | 50 | Not recorded | 68fa77f61e209131cea763ae |
| Llama3.2-Vision-11B | 50 | Not recorded | 68fa77f61e209131cea763ae |
| Gen3DEval w/ Fit3D | 44 | Not recorded | 68fa77f61e209131cea763ae |
| BLIP | 13 | Not recorded | 68fa77f61e209131cea763ae |
| PaliGemma | 10 | Not recorded | 68fa77f61e209131cea763ae |

### Gen3DEval (CLIP)

Accuracy: 86. Result ID: 6995775a3badfaffe66dd033.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Implementation](<https://shalini-maiti.github.io/gen3deval.github.io/>)

Recorded evaluation configuration: {"backbone":"CLIP"}. This record does not assert independent reproduction.

### Image Reward Score

Accuracy: 85. Result ID: 6995775a3badfaffe66dd02a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa719f1e209131cea52222.md>)

[Implementation](<https://github.com/THUDM/ImageReward>)

Recorded evaluation configuration: {"Prompt/Input handling":"None (empty strings passed)"}. This record does not assert independent reproduction.

### PickScore

Accuracy: 85. Result ID: 6995775a3badfaffe66dd02b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa713e1e209131cea500c1.md>)

[Implementation](<https://github.com/yuvalkirstain/PickScore>)

Recorded evaluation configuration: {"Prompt/Input handling":"None (empty strings passed)"}. This record does not assert independent reproduction.

### CLIP Score

Accuracy: 80. Result ID: 6995775a3badfaffe66dd029.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Prompt/Input handling":"None (empty strings passed)"}. This record does not assert independent reproduction.

### Gen3DEval (CLIP + DinoV2)

Accuracy: 74. Result ID: 6995775a3badfaffe66dd035.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Implementation](<https://shalini-maiti.github.io/gen3deval.github.io/>)

Recorded evaluation configuration: {"backbone":"CLIP + DinoV2"}. This record does not assert independent reproduction.

### Phi-3.5-Vision

Accuracy: 65. Result ID: 6995775a3badfaffe66dd02c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74ae1e209131cea632e1.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### LLaVA-Qwen-7B

Accuracy: 58. Result ID: 6995775a3badfaffe66dd02d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

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

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

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Gen3DEval w/ DinoV2

Accuracy: 58. Result ID: 6995775a3badfaffe66dd037.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Implementation](<https://shalini-maiti.github.io/gen3deval.github.io/>)

Recorded evaluation configuration: {"backbone":"DinoV2"}. This record does not assert independent reproduction.

### GPT-4o

Accuracy: 55. Result ID: 6995775a3badfaffe66dd031.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727a1e209131cea56a1e.md>)

[Implementation](<https://github.com/openai/evals>)

Recorded evaluation configuration: {"multi-image support":false}. This record does not assert independent reproduction.

### Gen3DEval (CLIP + Fit3D)

Accuracy: 53. Result ID: 6995775a3badfaffe66dd034.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Implementation](<https://shalini-maiti.github.io/gen3deval.github.io/>)

Recorded evaluation configuration: {"backbone":"CLIP + Fit3D"}. This record does not assert independent reproduction.

### LLaVA-Llama3-8b

Accuracy: 50. Result ID: 6995775a3badfaffe66dd02e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

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

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

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Llama3.2-Vision-11B

Accuracy: 50. Result ID: 6995775a3badfaffe66dd02f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"multi-image support":false}. This record does not assert independent reproduction.

### Gen3DEval w/ Fit3D

Accuracy: 44. Result ID: 6995775a3badfaffe66dd036.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Implementation](<https://shalini-maiti.github.io/gen3deval.github.io/>)

Recorded evaluation configuration: {"backbone":"Fit3D"}. This record does not assert independent reproduction.

### BLIP

Accuracy: 13. Result ID: 6995775a3badfaffe66dd030.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bd81e209131cea30874.md>)

[Implementation](<https://github.com/salesforce/BLIP>)

Recorded evaluation configuration: {"multi-image support":false}. This record does not assert independent reproduction.

### PaliGemma

Accuracy: 10. Result ID: 6995775a3badfaffe66dd032.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77f61e209131cea763ae.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa751e1e209131cea65e0c.md>)

Recorded evaluation configuration: {"multi-image support":false}. This record does not assert independent reproduction.

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