# Transformer Language Models without Positional Encodings Still Learn Positional Information

> Research article (Findings of the Association for Computational Linguistics: EMNLP 2022, 2022) · cited 50× · AI/ML

**Wikidata**: [openalex:W4385573607](https://www.wikidata.org/wiki/openalex:W4385573607)  
**Source**: https://4ort.xyz/entity/transformer-language-models-without-positional-encodings-still-learn-positional-information