# Visual Question Answering on SHAPES (test)

Source: [Visual Question Answering on SHAPES (test)](<https://www.sota2.com/research/sota/visual-question-answering-on-shapes-test>)

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

Data updated: 2026-02-26T05:34:33.602Z

Evaluation of model performance on the test split of the SHAPES dataset for Visual Question Answering, reporting accuracy broken down by object size categories.

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 | Visual Question Answering |
| Dataset (including any stated split) | SHAPES (test) |
| Metric | Accuracy (size 4) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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 (size 4)](<https://www.sota2.com/research/sota/visual-question-answering-on-shapes-test.md?metric=Accuracy%20(size%204)>): Higher is better; unit: PERCENT.

- [Accuracy (size 5)](<https://www.sota2.com/research/sota/visual-question-answering-on-shapes-test.md?metric=Accuracy%20(size%205)>): Higher is better; unit: PERCENT.

- [Accuracy (size 6)](<https://www.sota2.com/research/sota/visual-question-answering-on-shapes-test.md?metric=Accuracy%20(size%206)>): Higher is better; unit: PERCENT.

- [Overall Accuracy](<https://www.sota2.com/research/sota/visual-question-answering-on-shapes-test.md?metric=Overall%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (size 4) | Result date | Report paper |
| --- | --- | --- | --- |
| NMN (easy) | 97.7 | Not recorded | 68fa63801e209131ce9fef6e |
| NMN | 89.7 | Not recorded | 68fa63801e209131ce9fef6e |
| VIS+LSTM | 71.9 | Not recorded | 68fa63801e209131ce9fef6e |
| Majority | 64.4 | Not recorded | 68fa63801e209131ce9fef6e |

### NMN (easy)

Accuracy (size 4): 97.7. Result ID: 698bf62df729bc7f68e6e8e0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63801e209131ce9fef6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63801e209131ce9fef6e.md>)

Recorded evaluation configuration: {"backbone":"LeNet","training\_set":"modified (no size-6 questions)"}. This record does not assert independent reproduction.

### NMN

Accuracy (size 4): 89.7. Result ID: 698bf62df729bc7f68e6e8df.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63801e209131ce9fef6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63801e209131ce9fef6e.md>)

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

### VIS+LSTM

Accuracy (size 4): 71.9. Result ID: 698bf62df729bc7f68e6e8de.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63801e209131ce9fef6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62301e209131ce9f8146.md>)

[Implementation](<https://github.com/renmengye/imageqa-public>)

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

### Majority

Accuracy (size 4): 64.4. Result ID: 698bf62df729bc7f68e6e8dd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa63801e209131ce9fef6e.md>)

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

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