# Prompt Classification on ToxicChat Text Prompt

Source: [Prompt Classification on ToxicChat Text Prompt](<https://www.sota2.com/research/sota/prompt-classification-on-toxicchat-text-prompt>)

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

Data updated: 2026-02-27T11:29:59.993Z

Evaluation of model performance on classifying prompts using the ToxicChat Text Prompt dataset, typically measured using F1 score.

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 | Prompt Classification |
| Dataset (including any stated split) | ToxicChat Text Prompt |
| Metric | F1 Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| 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

- [F1 Score](<https://www.sota2.com/research/sota/prompt-classification-on-toxicchat-text-prompt.md?metric=F1%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | F1 Score | Result date | Report paper |
| --- | --- | --- | --- |
| GPT4o-mini | 96.27 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| ProGuard-7B | 96.07 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| GuardReasoner-8B | 95.8 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| Gemini2.5-Flash | 95.43 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| GuardReasonerVL-7B | 95.32 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| ProGuard-3B | 95.03 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| LlamaGuard-7B | 94.53 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| ShieldGemma-9B | 93.97 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| GPT-OSS-SafeGuard-20B | 93.72 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| LlamaGuard2-8B | 92.33 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| LlamaGuard3-11B-Vision | 92.05 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| LlamaGuard3-8B | 91.56 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| LlamaGuard4-12B | 91.15 | Not recorded | 69535c9b53fdd5c36173dbb6 |
| WildGuard-7B | 88.99 | Not recorded | 69535c9b53fdd5c36173dbb6 |

### GPT4o-mini

F1 Score: 96.27. Result ID: 69a1802b35dfcabb00e774a3.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75681e209131cea67a98.md>)

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

### ProGuard-7B

F1 Score: 96.07. Result ID: 69a1802b35dfcabb00e774a1.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Implementation](<https://github.com/yushaohan/ProGuard>)

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

### GuardReasoner-8B

F1 Score: 95.8. Result ID: 69a1802b35dfcabb00e77497.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c491e209131cea8d665.md>)

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

### Gemini2.5-Flash

F1 Score: 95.43. Result ID: 69a1802b35dfcabb00e774a2.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c311e209131cea8cef5.md>)

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

### GuardReasonerVL-7B

F1 Score: 95.32. Result ID: 69a1802b35dfcabb00e7749d.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa789e1e209131cea7963f.md>)

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

### ProGuard-3B

F1 Score: 95.03. Result ID: 69a1802b35dfcabb00e774a0.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Implementation](<https://github.com/yushaohan/ProGuard>)

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

### LlamaGuard-7B

F1 Score: 94.53. Result ID: 69a1802b35dfcabb00e77498.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

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

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

### ShieldGemma-9B

F1 Score: 93.97. Result ID: 69a1802b35dfcabb00e7749b.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74661e209131cea617ba.md>)

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

### GPT-OSS-SafeGuard-20B

F1 Score: 93.72. Result ID: 69a1802b35dfcabb00e77496.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

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

### LlamaGuard2-8B

F1 Score: 92.33. Result ID: 69a1802b35dfcabb00e77499.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

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

### LlamaGuard3-11B-Vision

F1 Score: 92.05. Result ID: 69a1802b35dfcabb00e7749e.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75c21e209131cea69c6e.md>)

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

### LlamaGuard3-8B

F1 Score: 91.56. Result ID: 69a1802b35dfcabb00e7749a.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

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

### LlamaGuard4-12B

F1 Score: 91.15. Result ID: 69a1802b35dfcabb00e7749f.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

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

### WildGuard-7B

F1 Score: 88.99. Result ID: 69a1802b35dfcabb00e7749c.

[Reporting paper](<https://www.sota2.com/research/paper/69535c9b53fdd5c36173dbb6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76291e209131cea6b9a6.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.
