# POS Tagging on Universal Dependencies (UD)

Source: [POS Tagging on Universal Dependencies (UD)](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud>)

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

Data updated: 2026-02-26T05:22:30.494Z

Evaluation of Part-of-Speech (POS) Tagging performance across various languages within the Universal Dependencies (UD) corpus.

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 | POS Tagging |
| Dataset (including any stated split) | Universal Dependencies (UD) |
| Metric | AMH Accuracy |
| Unit | PERCENT |
| 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

- [AMH Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=AMH%20Accuracy>): Higher is better; unit: PERCENT.

- [Lij Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=Lij%20Accuracy>): Higher is better; unit: PERCENT.

- [Olo Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=Olo%20Accuracy>): Higher is better; unit: PERCENT.

- [San Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=San%20Accuracy>): Higher is better; unit: PERCENT.

- [Snd Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=Snd%20Accuracy>): Higher is better; unit: PERCENT.

- [Sin Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=Sin%20Accuracy>): Higher is better; unit: PERCENT.

- [Tam Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=Tam%20Accuracy>): Higher is better; unit: PERCENT.

- [Tgl Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=Tgl%20Accuracy>): Higher is better; unit: PERCENT.

- [Tat Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=Tat%20Accuracy>): Higher is better; unit: PERCENT.

- [Average Accuracy](<https://www.sota2.com/research/sota/pos-tagging-on-universal-dependencies-ud.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AMH Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| MAD-X (XLM-R) | 47.72 | Not recorded | 68fa748d1e209131cea626ca |
| XLM-R | 46.02 | Not recorded | 68fa748d1e209131cea626ca |
| UniBridge (XLM-R) | 40.88 | Not recorded | 68fa748d1e209131cea626ca |

### MAD-X (XLM-R)

AMH Accuracy: 47.72. Result ID: 69934b3e3a6c31484914fbc7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa748d1e209131cea626ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa686b1e209131cea1c005.md>)

[Implementation](<https://github.com/Adapter-Hub/adapter-transformers>)

Recorded evaluation configuration: {"Backbone":"XLM-R"}. This record does not assert independent reproduction.

### XLM-R

AMH Accuracy: 46.02. Result ID: 69934b3e3a6c31484914fbc6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa748d1e209131cea626ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67521e209131cea15706.md>)

Recorded evaluation configuration: {"Backbone":"XLM-R"}. This record does not assert independent reproduction.

### UniBridge (XLM-R)

AMH Accuracy: 40.88. Result ID: 69934b3e3a6c31484914fbc8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa748d1e209131cea626ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa748d1e209131cea626ca.md>)

[Implementation](<https://github.com/TokisakiKurumi2001/UniBridge>)

Recorded evaluation configuration: {"Backbone":"XLM-R"}. This record does not assert independent reproduction.

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