# Counterfactual Hallucination Detection on ROCO v2

Source: [Counterfactual Hallucination Detection on ROCO v2](<https://www.sota2.com/research/sota/counterfactual-hallucination-detection-on-roco-v2>)

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

Data updated: 2026-05-08T09:57:54.924Z

Evaluation of model performance on counterfactual hallucination detection using the ROCO v2 dataset, reporting Pairwise 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 | Counterfactual Hallucination Detection |
| Dataset (including any stated split) | ROCO v2 |
| Metric | Pairwise Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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

- [Pairwise Accuracy](<https://www.sota2.com/research/sota/counterfactual-hallucination-detection-on-roco-v2.md?metric=Pairwise%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Pairwise Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Soft-MSD | 74.9 | Not recorded | 69fd6e145497b3b670c89401 |
| MSD | 73.3 | Not recorded | 69fd6e145497b3b670c89401 |
| CLIPScore | 71.7 | Not recorded | 69fd6e145497b3b670c89401 |

### Soft-MSD

Pairwise Accuracy: 74.9. Result ID: 69fdb39747f34ce924e12122.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e145497b3b670c89401.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e145497b3b670c89401.md>)

Recorded evaluation configuration: {"Backbone":"BiomedCLIP \[44\]"}. This record does not assert independent reproduction.

### MSD

Pairwise Accuracy: 73.3. Result ID: 69fdb39747f34ce924e12121.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e145497b3b670c89401.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69fd6e145497b3b670c89401.md>)

Recorded evaluation configuration: {"Backbone":"BiomedCLIP \[44\]"}. This record does not assert independent reproduction.

### CLIPScore

Pairwise Accuracy: 71.7. Result ID: 69fdb39747f34ce924e12120.

[Reporting paper](<https://www.sota2.com/research/paper/69fd6e145497b3b670c89401.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c0f1e209131cea31cc9.md>)

Recorded evaluation configuration: {"Backbone":"BiomedCLIP \[44\]"}. This record does not assert independent reproduction.

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