# Error Detection on TruthfulQA 200-question subset

Source: [Error Detection on TruthfulQA 200-question subset](<https://www.sota2.com/research/sota/error-detection-on-truthfulqa-200-question-subset>)

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

Data updated: 2026-03-26T14:25:31.551Z

Evaluation of model performance on the specific 200-question subset of the TruthfulQA dataset, focusing on error detection capabilities using metrics like AUROC.

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 | Error Detection |
| Dataset (including any stated split) | TruthfulQA 200-question subset |
| Metric | AUROC |
| Unit | SCALAR |
| 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

- [AUROC](<https://www.sota2.com/research/sota/error-detection-on-truthfulqa-200-question-subset.md?metric=AUROC>): Higher is better; unit: SCALAR.

- [95% CI](<https://www.sota2.com/research/sota/error-detection-on-truthfulqa-200-question-subset.md?metric=95%25%20CI>): Higher is better; unit: SCALAR.

- [Cost (NLI Steps)](<https://www.sota2.com/research/sota/error-detection-on-truthfulqa-200-question-subset.md?metric=Cost%20(NLI%20Steps)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | AUROC | Result date | Report paper |
| --- | --- | --- | --- |
| SE-NLI | 0.512 | Not recorded | 69c4bdb60636e5bf4afc7760 |
| SE-NLI | 0.511 | Not recorded | 69c4bdb60636e5bf4afc7760 |
| SE-NLI | 0.501 | Not recorded | 69c4bdb60636e5bf4afc7760 |

### SE-NLI

AUROC: 0.512. Result ID: 69c541bf47f34ce924af22fc.

[Reporting paper](<https://www.sota2.com/research/paper/69c4bdb60636e5bf4afc7760.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f141e209131cea43bf3.md>)

Recorded evaluation configuration: {"Backbone":"DeBERTa-base","Sample size (N)":"10"}. This record does not assert independent reproduction.

### SE-NLI

AUROC: 0.511. Result ID: 69c541bf47f34ce924af22fb.

[Reporting paper](<https://www.sota2.com/research/paper/69c4bdb60636e5bf4afc7760.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f141e209131cea43bf3.md>)

Recorded evaluation configuration: {"Backbone":"DeBERTa-large","Sample size (N)":"10"}. This record does not assert independent reproduction.

### SE-NLI

AUROC: 0.501. Result ID: 69c541bf47f34ce924af22fd.

[Reporting paper](<https://www.sota2.com/research/paper/69c4bdb60636e5bf4afc7760.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f141e209131cea43bf3.md>)

Recorded evaluation configuration: {"Backbone":"DeBERTa-xsmall","Sample size (N)":"10"}. This record does not assert independent reproduction.

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