# Harmful Content Detection on FoodGuardBench (test)

Source: [Harmful Content Detection on FoodGuardBench (test)](<https://www.sota2.com/research/sota/harmful-content-detection-on-foodguardbench-test>)

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

Data updated: 2026-04-03T07:24:07.951Z

Evaluation of model performance on the FoodGuardBench (test) dataset for the task of Harmful Content Detection, reporting metrics such as FNR, FPR, F1, and ACC.

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 | Harmful Content Detection |
| Dataset (including any stated split) | FoodGuardBench (test) |
| Metric | FNR |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 7 of 7 |
| 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

- [FNR](<https://www.sota2.com/research/sota/harmful-content-detection-on-foodguardbench-test.md?metric=FNR>): Lower is better; unit: PERCENT.

- [FPR](<https://www.sota2.com/research/sota/harmful-content-detection-on-foodguardbench-test.md?metric=FPR>): Lower is better; unit: PERCENT.

- [F1 Score](<https://www.sota2.com/research/sota/harmful-content-detection-on-foodguardbench-test.md?metric=F1%20Score>): Higher is better; unit: PERCENT.

- [Accuracy](<https://www.sota2.com/research/sota/harmful-content-detection-on-foodguardbench-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | FNR | Result date | Report paper |
| --- | --- | --- | --- |
| FoodGuard-4B | 2.75 | Not recorded | 69cf496d0636e5bf4afcd900 |
| Qwen3Guard-4B | 21.43 | Not recorded | 69cf496d0636e5bf4afcd900 |
| Qwen3-8B | 24.55 | Not recorded | 69cf496d0636e5bf4afcd900 |
| LLaMA3.1-8B | 26.25 | Not recorded | 69cf496d0636e5bf4afcd900 |
| Qwen3Guard-8B | 29.52 | Not recorded | 69cf496d0636e5bf4afcd900 |
| LLaMA-Guard3-8B | 42.4 | Not recorded | 69cf496d0636e5bf4afcd900 |
| LLaMA-Guard4-12B | 59.71 | Not recorded | 69cf496d0636e5bf4afcd900 |

### FoodGuard-4B

FNR: 2.75. Result ID: 69cf6a8847f34ce924b8dba0.

[Reporting paper](<https://www.sota2.com/research/paper/69cf496d0636e5bf4afcd900.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69cf496d0636e5bf4afcd900.md>)

Recorded evaluation configuration: {"Parameters":"4B"}. This record does not assert independent reproduction.

### Qwen3Guard-4B

FNR: 21.43. Result ID: 69cf6a8847f34ce924b8db9f.

[Reporting paper](<https://www.sota2.com/research/paper/69cf496d0636e5bf4afcd900.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c351e209131cea8d09c.md>)

Recorded evaluation configuration: {"Parameters":"4B"}. This record does not assert independent reproduction.

### Qwen3-8B

FNR: 24.55. Result ID: 69cf6a8847f34ce924b8db9b.

[Reporting paper](<https://www.sota2.com/research/paper/69cf496d0636e5bf4afcd900.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa788e1e209131cea79116.md>)

Recorded evaluation configuration: {"Parameters":"8B"}. This record does not assert independent reproduction.

### LLaMA3.1-8B

FNR: 26.25. Result ID: 69cf6a8847f34ce924b8db9a.

[Reporting paper](<https://www.sota2.com/research/paper/69cf496d0636e5bf4afcd900.md>)

Recorded evaluation configuration: {"Parameters":"8B"}. This record does not assert independent reproduction.

### Qwen3Guard-8B

FNR: 29.52. Result ID: 69cf6a8847f34ce924b8db9e.

[Reporting paper](<https://www.sota2.com/research/paper/69cf496d0636e5bf4afcd900.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c351e209131cea8d09c.md>)

Recorded evaluation configuration: {"Parameters":"8B"}. This record does not assert independent reproduction.

### LLaMA-Guard3-8B

FNR: 42.4. Result ID: 69cf6a8847f34ce924b8db9d.

[Reporting paper](<https://www.sota2.com/research/paper/69cf496d0636e5bf4afcd900.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Parameters":"8B"}. This record does not assert independent reproduction.

### LLaMA-Guard4-12B

FNR: 59.71. Result ID: 69cf6a8847f34ce924b8db9c.

[Reporting paper](<https://www.sota2.com/research/paper/69cf496d0636e5bf4afcd900.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Parameters":"12B"}. This record does not assert independent reproduction.

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