# Unsafe Prompt Detection on ToxicChat

Source: [Unsafe Prompt Detection on ToxicChat](<https://www.sota2.com/research/sota/unsafe-prompt-detection-on-toxicchat>)

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

Data updated: 2026-02-26T05:25:39.555Z

Evaluation of model performance on detecting unsafe prompts using the ToxicChat dataset, primarily measured by Area Under the Precision-Recall Curve (AUPRC).

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 | Unsafe prompt detection |
| Dataset (including any stated split) | ToxicChat |
| Metric | AUPRC |
| 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

- [AUPRC](<https://www.sota2.com/research/sota/unsafe-prompt-detection-on-toxicchat.md?metric=AUPRC>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AUPRC | Result date | Report paper |
| --- | --- | --- | --- |
| GradSafe-Zero | 75.5 | Not recorded | 68fa736c1e209131cea5c676 |
| Llama Guard | 63.5 | Not recorded | 68fa736c1e209131cea5c676 |
| OpenAI Moderation API | 60.4 | Not recorded | 68fa736c1e209131cea5c676 |
| Perspective API | 48.7 | Not recorded | 68fa736c1e209131cea5c676 |

### GradSafe-Zero

AUPRC: 75.5. Result ID: 6992ea2a3a6c314849125617.

[Reporting paper](<https://www.sota2.com/research/paper/68fa736c1e209131cea5c676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa736c1e209131cea5c676.md>)

[Implementation](<https://github.com/xyq7/GradSafe>)

Recorded evaluation configuration: {"Backbone":"Llama-2-7b-chat-hf"}. This record does not assert independent reproduction.

### Llama Guard

AUPRC: 63.5. Result ID: 6992ea2a3a6c314849125616.

[Reporting paper](<https://www.sota2.com/research/paper/68fa736c1e209131cea5c676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa717b1e209131cea5186d.md>)

Recorded evaluation configuration: {"Backbone":"Llama-2 7b"}. This record does not assert independent reproduction.

### OpenAI Moderation API

AUPRC: 60.4. Result ID: 6992ea2a3a6c314849125614.

[Reporting paper](<https://www.sota2.com/research/paper/68fa736c1e209131cea5c676.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e8b1e209131cea40729.md>)

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

### Perspective API

AUPRC: 48.7. Result ID: 6992ea2a3a6c314849125615.

[Reporting paper](<https://www.sota2.com/research/paper/68fa736c1e209131cea5c676.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.
