# Hallucination Detection on VQA-Med Open-Ended 2019 (test)

Source: [Hallucination Detection on VQA-Med Open-Ended 2019 (test)](<https://www.sota2.com/research/sota/hallucination-detection-on-vqa-med-open-ended-2019-test>)

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

Data updated: 2026-02-26T05:24:56.363Z

Evaluation of model performance on hallucination detection using the test split of the VQA-Med Open-Ended 2019 dataset, measured by AUC and AUG.

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 | Hallucination Detection |
| Dataset (including any stated split) | VQA-Med Open-Ended 2019 (test) |
| Metric | AUC |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 18 of 18 |
| 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

- [AUC](<https://www.sota2.com/research/sota/hallucination-detection-on-vqa-med-open-ended-2019-test.md?metric=AUC>): Higher is better; unit: PERCENT.

- [AUG](<https://www.sota2.com/research/sota/hallucination-detection-on-vqa-med-open-ended-2019-test.md?metric=AUG>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AUC | Result date | Report paper |
| --- | --- | --- | --- |
| VASE + V-Loop | 78.14 | Not recorded | 697846caa2b28e761d87c710 |
| RadFlag + V-Loop | 76.14 | Not recorded | 697846caa2b28e761d87c710 |
| VASE + V-Loop | 76.01 | Not recorded | 697846caa2b28e761d87c710 |
| SE + V-Loop | 75.6 | Not recorded | 697846caa2b28e761d87c710 |
| VASE + V-Loop | 73.51 | Not recorded | 697846caa2b28e761d87c710 |
| RadFlag + V-Loop | 73.42 | Not recorded | 697846caa2b28e761d87c710 |
| SE + V-Loop | 72.51 | Not recorded | 697846caa2b28e761d87c710 |
| VASE | 72.35 | Not recorded | 697846caa2b28e761d87c710 |
| RadFlag + V-Loop | 71.11 | Not recorded | 697846caa2b28e761d87c710 |
| VASE | 70.92 | Not recorded | 697846caa2b28e761d87c710 |
| SE + V-Loop | 69.02 | Not recorded | 697846caa2b28e761d87c710 |
| VASE | 68.02 | Not recorded | 697846caa2b28e761d87c710 |
| RadFlag | 67.69 | Not recorded | 697846caa2b28e761d87c710 |
| SE | 66.18 | Not recorded | 697846caa2b28e761d87c710 |
| RadFlag | 64.91 | Not recorded | 697846caa2b28e761d87c710 |
| SE | 61.88 | Not recorded | 697846caa2b28e761d87c710 |
| RadFlag | 55.24 | Not recorded | 697846caa2b28e761d87c710 |
| SE | 51.8 | Not recorded | 697846caa2b28e761d87c710 |

### VASE + V-Loop

AUC: 78.14. Result ID: 69978d685698665ec4e29316.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"InternVL3-8B \[20\]"}. This record does not assert independent reproduction.

### RadFlag + V-Loop

AUC: 76.14. Result ID: 69978d685698665ec4e29310.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"InternVL3-8B \[20\]"}. This record does not assert independent reproduction.

### VASE + V-Loop

AUC: 76.01. Result ID: 69978d685698665ec4e29314.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"Lingshu-7B \[40\]"}. This record does not assert independent reproduction.

### SE + V-Loop

AUC: 75.6. Result ID: 69978d685698665ec4e2930a.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"InternVL3-8B \[20\]"}. This record does not assert independent reproduction.

### VASE + V-Loop

AUC: 73.51. Result ID: 69978d685698665ec4e29312.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"MedGemma-4B-it \[34\]"}. This record does not assert independent reproduction.

### RadFlag + V-Loop

AUC: 73.42. Result ID: 69978d685698665ec4e2930e.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"Lingshu-7B \[40\]"}. This record does not assert independent reproduction.

### SE + V-Loop

AUC: 72.51. Result ID: 69978d685698665ec4e29308.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"Lingshu-7B \[40\]"}. This record does not assert independent reproduction.

### VASE

AUC: 72.35. Result ID: 69978d685698665ec4e29315.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"InternVL3-8B \[20\]"}. This record does not assert independent reproduction.

### RadFlag + V-Loop

AUC: 71.11. Result ID: 69978d685698665ec4e2930c.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"MedGemma-4B-it \[34\]"}. This record does not assert independent reproduction.

### VASE

AUC: 70.92. Result ID: 69978d685698665ec4e29311.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"MedGemma-4B-it \[34\]"}. This record does not assert independent reproduction.

### SE + V-Loop

AUC: 69.02. Result ID: 69978d685698665ec4e29306.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

[Method origin paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"MedGemma-4B-it \[34\]"}. This record does not assert independent reproduction.

### VASE

AUC: 68.02. Result ID: 69978d685698665ec4e29313.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"Lingshu-7B \[40\]"}. This record does not assert independent reproduction.

### RadFlag

AUC: 67.69. Result ID: 69978d685698665ec4e2930f.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"InternVL3-8B \[20\]"}. This record does not assert independent reproduction.

### SE

AUC: 66.18. Result ID: 69978d685698665ec4e29309.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"InternVL3-8B \[20\]"}. This record does not assert independent reproduction.

### RadFlag

AUC: 64.91. Result ID: 69978d685698665ec4e2930b.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"MedGemma-4B-it \[34\]"}. This record does not assert independent reproduction.

### SE

AUC: 61.88. Result ID: 69978d685698665ec4e29305.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"MedGemma-4B-it \[34\]"}. This record does not assert independent reproduction.

### RadFlag

AUC: 55.24. Result ID: 69978d685698665ec4e2930d.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"Lingshu-7B \[40\]"}. This record does not assert independent reproduction.

### SE

AUC: 51.8. Result ID: 69978d685698665ec4e29307.

[Reporting paper](<https://www.sota2.com/research/paper/697846caa2b28e761d87c710.md>)

Recorded evaluation configuration: {"Backbone":"Lingshu-7B \[40\]"}. This record does not assert independent reproduction.

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