# Toxicity Classification on Multilingual Jigsaw

Source: [Toxicity Classification on Multilingual Jigsaw](<https://www.sota2.com/research/sota/toxicity-classification-on-multilingual-jigsaw>)

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

Data updated: 2026-02-26T05:45:33.832Z

Evaluation of model performance on the Multilingual Jigsaw dataset for the task of Toxicity Classification, reporting scores across multiple languages (French, Portuguese, Russian, Spanish, Turkish).

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 | Toxicity Classification |
| Dataset (including any stated split) | Multilingual Jigsaw |
| Metric | French Accuracy |
| 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

- [French Accuracy](<https://www.sota2.com/research/sota/toxicity-classification-on-multilingual-jigsaw.md?metric=French%20Accuracy>): Higher is better; unit: PERCENT.

- [Portuguese Accuracy](<https://www.sota2.com/research/sota/toxicity-classification-on-multilingual-jigsaw.md?metric=Portuguese%20Accuracy>): Higher is better; unit: PERCENT.

- [Russian Accuracy](<https://www.sota2.com/research/sota/toxicity-classification-on-multilingual-jigsaw.md?metric=Russian%20Accuracy>): Higher is better; unit: PERCENT.

- [Spanish Accuracy](<https://www.sota2.com/research/sota/toxicity-classification-on-multilingual-jigsaw.md?metric=Spanish%20Accuracy>): Higher is better; unit: PERCENT.

- [Turkish Accuracy](<https://www.sota2.com/research/sota/toxicity-classification-on-multilingual-jigsaw.md?metric=Turkish%20Accuracy>): Higher is better; unit: PERCENT.

- [Non-English Accuracy](<https://www.sota2.com/research/sota/toxicity-classification-on-multilingual-jigsaw.md?metric=Non-English%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | French Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| PaLM 2 | 87.94 | Not recorded | 68fa708e1e209131cea4bc0a |
| PaLM | 86.16 | Not recorded | 68fa708e1e209131cea4bc0a |
| PaLM 2 | 84.11 | Not recorded | 68fa708e1e209131cea4bc0a |
| PaLM | 78.99 | Not recorded | 68fa708e1e209131cea4bc0a |

### PaLM 2

French Accuracy: 87.94. Result ID: 699148263a6c3148490700e7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa708e1e209131cea4bc0a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa708e1e209131cea4bc0a.md>)

Recorded evaluation configuration: {"number of shots":"10-shot"}. This record does not assert independent reproduction.

### PaLM

French Accuracy: 86.16. Result ID: 699148263a6c3148490700e6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa708e1e209131cea4bc0a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d931e209131cea3abf2.md>)

Recorded evaluation configuration: {"number of shots":"10-shot"}. This record does not assert independent reproduction.

### PaLM 2

French Accuracy: 84.11. Result ID: 699148263a6c3148490700e5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa708e1e209131cea4bc0a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa708e1e209131cea4bc0a.md>)

Recorded evaluation configuration: {"number of shots":"0-shot"}. This record does not assert independent reproduction.

### PaLM

French Accuracy: 78.99. Result ID: 699148263a6c3148490700e4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa708e1e209131cea4bc0a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d931e209131cea3abf2.md>)

Recorded evaluation configuration: {"number of shots":"0-shot"}. This record does not assert independent reproduction.

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