# Knowledge-Intensive Visual Question Answering on Museum Artifact Dataset V+K 1.0 (test)

Source: [Knowledge-Intensive Visual Question Answering on Museum Artifact Dataset V+K 1.0 (test)](<https://www.sota2.com/research/sota/knowledge-intensive-visual-question-answering-on-museum-artifact-dataset-v-plus-k-1-0-test>)

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

Data updated: 2026-07-08T07:52:01.979Z

Evaluation of model performance on the test split of the Museum Artifact Dataset V+K 1.0 for Knowledge-Intensive Visual Question Answering.

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 | Knowledge-Intensive Visual Question Answering |
| Dataset (including any stated split) | Museum Artifact Dataset V+K 1.0 (test) |
| Metric | Answer Accuracy |
| 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

- [Answer Accuracy](<https://www.sota2.com/research/sota/knowledge-intensive-visual-question-answering-on-museum-artifact-dataset-v-plus-k-1-0-test.md?metric=Answer%20Accuracy>): Higher is better; unit: PERCENT.

- [Hallucination Rate](<https://www.sota2.com/research/sota/knowledge-intensive-visual-question-answering-on-museum-artifact-dataset-v-plus-k-1-0-test.md?metric=Hallucination%20Rate>): Lower is better; unit: PERCENT.

- [Groundedness Score](<https://www.sota2.com/research/sota/knowledge-intensive-visual-question-answering-on-museum-artifact-dataset-v-plus-k-1-0-test.md?metric=Groundedness%20Score>): Higher is better; unit: PERCENT.

- [Neutrality Score](<https://www.sota2.com/research/sota/knowledge-intensive-visual-question-answering-on-museum-artifact-dataset-v-plus-k-1-0-test.md?metric=Neutrality%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Answer Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen3-VL-8B (Direct) | 69 | Not recorded | 6a4dd9d44da6617fe9198db4 |
| Qwen3-VL-8B + Search | 59 | Not recorded | 6a4dd9d44da6617fe9198db4 |
| VaseAgent (Full) | 54 | Not recorded | 6a4dd9d44da6617fe9198db4 |
| VaseAgent (w/o GRPO) | 53 | Not recorded | 6a4dd9d44da6617fe9198db4 |

### Qwen3-VL-8B (Direct)

Answer Accuracy: 69. Result ID: 6a4e019260b88bb339d7883a.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9d44da6617fe9198db4.md>)

Recorded evaluation configuration: {"search":"disabled","reliability control":"none"}. This record does not assert independent reproduction.

### Qwen3-VL-8B + Search

Answer Accuracy: 59. Result ID: 6a4e019260b88bb339d7883b.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9d44da6617fe9198db4.md>)

Recorded evaluation configuration: {"search":"enabled","reliability control":"none"}. This record does not assert independent reproduction.

### VaseAgent (Full)

Answer Accuracy: 54. Result ID: 6a4e019260b88bb339d7883d.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9d44da6617fe9198db4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a4dd9d44da6617fe9198db4.md>)

[Implementation](<https://github.com/AIGeeksGroup/VaseMuseum>)

Recorded evaluation configuration: {"search":"enabled","reliability control":"complete system","GRPO selection":"enabled"}. This record does not assert independent reproduction.

### VaseAgent (w/o GRPO)

Answer Accuracy: 53. Result ID: 6a4e019260b88bb339d7883c.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9d44da6617fe9198db4.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a4dd9d44da6617fe9198db4.md>)

[Implementation](<https://github.com/AIGeeksGroup/VaseMuseum>)

Recorded evaluation configuration: {"search":"enabled","reliability control":"source/response control","GRPO selection":"disabled"}. This record does not assert independent reproduction.

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