UniBERT: Adversarial Training for Language-Universal Representations

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Avram, Andrei-Marius, Lupaşcu, Marian, Cercel, Dumitru-Clementin, Mironică, Ionuţ, Trăuşan-Matu, Ştefan
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916929254457344
author Avram, Andrei-Marius
Lupaşcu, Marian
Cercel, Dumitru-Clementin
Mironică, Ionuţ
Trăuşan-Matu, Ştefan
author_facet Avram, Andrei-Marius
Lupaşcu, Marian
Cercel, Dumitru-Clementin
Mironică, Ionuţ
Trăuşan-Matu, Ştefan
contents This paper presents UniBERT, a compact multilingual language model that uses an innovative training framework that integrates three components: masked language modeling, adversarial training, and knowledge distillation. Pre-trained on a meticulously curated Wikipedia corpus spanning 107 languages, UniBERT is designed to reduce the computational demands of large-scale models while maintaining competitive performance across various natural language processing tasks. Comprehensive evaluations on four tasks - named entity recognition, natural language inference, question answering, and semantic textual similarity - demonstrate that our multilingual training strategy enhanced by an adversarial objective significantly improves cross-lingual generalization. Specifically, UniBERT models show an average relative improvement of 7.72% over traditional baselines, which achieved an average relative improvement of only 1.17%, and statistical analysis confirms the significance of these gains (p-value = 0.0181). This work highlights the benefits of combining adversarial training and knowledge distillation to build scalable and robust language models, thus advancing the field of multilingual and cross-lingual natural language processing.
format Preprint
id arxiv_https___arxiv_org_abs_2503_12608
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle UniBERT: Adversarial Training for Language-Universal Representations
Avram, Andrei-Marius
Lupaşcu, Marian
Cercel, Dumitru-Clementin
Mironică, Ionuţ
Trăuşan-Matu, Ştefan
Computation and Language
This paper presents UniBERT, a compact multilingual language model that uses an innovative training framework that integrates three components: masked language modeling, adversarial training, and knowledge distillation. Pre-trained on a meticulously curated Wikipedia corpus spanning 107 languages, UniBERT is designed to reduce the computational demands of large-scale models while maintaining competitive performance across various natural language processing tasks. Comprehensive evaluations on four tasks - named entity recognition, natural language inference, question answering, and semantic textual similarity - demonstrate that our multilingual training strategy enhanced by an adversarial objective significantly improves cross-lingual generalization. Specifically, UniBERT models show an average relative improvement of 7.72% over traditional baselines, which achieved an average relative improvement of only 1.17%, and statistical analysis confirms the significance of these gains (p-value = 0.0181). This work highlights the benefits of combining adversarial training and knowledge distillation to build scalable and robust language models, thus advancing the field of multilingual and cross-lingual natural language processing.
title UniBERT: Adversarial Training for Language-Universal Representations
topic Computation and Language
url https://arxiv.org/abs/2503.12608