# Referring Expression Comprehension on RefCOCO (Accuracy)

Source: [Referring Expression Comprehension on RefCOCO (Accuracy)](<https://www.sota2.com/research/sota/referring-expression-comprehension-on-refcoco-accuracy>)

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

Data updated: 2026-02-26T05:31:49.849Z

Evaluation of Referring Expression Comprehension performance on the RefCOCO dataset, measured using 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 | Referring Expression Comprehension |
| Dataset (including any stated split) | RefCOCO |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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/referring-expression-comprehension-on-refcoco-accuracy.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen2.5-VL-7B | 89.5 | Not recorded | 69802f6ba8b01e76cc8e14f5 |
| VisionTrim | 86.8 | Not recorded | 69802f6ba8b01e76cc8e14f5 |
| SparseVLM | 78.4 | Not recorded | 69802f6ba8b01e76cc8e14f5 |
| PaliGemma2-10B | 78.2 | Not recorded | 6944dc2b423ea42265729d1f |
| PaliGemma2-3B + AuditDM | 77.8 | Not recorded | 6944dc2b423ea42265729d1f |
| PDrop | 77.5 | Not recorded | 69802f6ba8b01e76cc8e14f5 |
| PaliGemma2-28B | 77.3 | Not recorded | 6944dc2b423ea42265729d1f |
| PaliGemma2-3B | 76.3 | Not recorded | 6944dc2b423ea42265729d1f |
| FastV | 73.6 | Not recorded | 69802f6ba8b01e76cc8e14f5 |

### Qwen2.5-VL-7B

Accuracy: 89.5. Result ID: 699773715698665ec4c77f68.

[Reporting paper](<https://www.sota2.com/research/paper/69802f6ba8b01e76cc8e14f5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"Token Budget":"Full"}. This record does not assert independent reproduction.

### VisionTrim

Accuracy: 86.8. Result ID: 699773715698665ec4c77f6c.

[Reporting paper](<https://www.sota2.com/research/paper/69802f6ba8b01e76cc8e14f5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69802f6ba8b01e76cc8e14f5.md>)

[Implementation](<https://github.com/hanxunyu/VisionTrim>)

Recorded evaluation configuration: {"Token Budget":"~1/3"}. This record does not assert independent reproduction.

### SparseVLM

Accuracy: 78.4. Result ID: 699773715698665ec4c77f6a.

[Reporting paper](<https://www.sota2.com/research/paper/69802f6ba8b01e76cc8e14f5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79181e209131cea7c1b1.md>)

Recorded evaluation configuration: {"Token Budget":"~1/3"}. This record does not assert independent reproduction.

### PaliGemma2-10B

Accuracy: 78.2. Result ID: 69982ce95698665ec487eb80.

[Reporting paper](<https://www.sota2.com/research/paper/6944dc2b423ea42265729d1f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76001e209131cea6ad66.md>)

Recorded evaluation configuration: {"Resolution":"448x448","Fine-tuning":"per-task"}. This record does not assert independent reproduction.

### PaliGemma2-3B + AuditDM

Accuracy: 77.8. Result ID: 69982ce95698665ec487eb83.

[Reporting paper](<https://www.sota2.com/research/paper/6944dc2b423ea42265729d1f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6944dc2b423ea42265729d1f.md>)

Recorded evaluation configuration: {"Resolution":"448x448","Fine-tuning":"per-task"}. This record does not assert independent reproduction.

### PDrop

Accuracy: 77.5. Result ID: 699773715698665ec4c77f6b.

[Reporting paper](<https://www.sota2.com/research/paper/69802f6ba8b01e76cc8e14f5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772c1e209131cea71934.md>)

Recorded evaluation configuration: {"Token Budget":"~1/3"}. This record does not assert independent reproduction.

### PaliGemma2-28B

Accuracy: 77.3. Result ID: 69982ce95698665ec487eb81.

[Reporting paper](<https://www.sota2.com/research/paper/6944dc2b423ea42265729d1f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76001e209131cea6ad66.md>)

Recorded evaluation configuration: {"Resolution":"448x448","Fine-tuning":"per-task"}. This record does not assert independent reproduction.

### PaliGemma2-3B

Accuracy: 76.3. Result ID: 69982ce95698665ec487eb82.

[Reporting paper](<https://www.sota2.com/research/paper/6944dc2b423ea42265729d1f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76001e209131cea6ad66.md>)

Recorded evaluation configuration: {"Resolution":"448x448","Fine-tuning":"per-task"}. This record does not assert independent reproduction.

### FastV

Accuracy: 73.6. Result ID: 699773715698665ec4c77f69.

[Reporting paper](<https://www.sota2.com/research/paper/69802f6ba8b01e76cc8e14f5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74ae1e209131cea632b7.md>)

Recorded evaluation configuration: {"Token Budget":"~1/3"}. This record does not assert independent reproduction.

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