Rethinking what Matters: Effective and Robust Multilingual Realignment for Low-Resource Languages
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915607439474688 |
|---|---|
| author | Nguyen, Quang Phuoc Anugraha, David Gaschi, Felix Cheng, Jun Bin Lee, En-Shiun Annie |
| author_facet | Nguyen, Quang Phuoc Anugraha, David Gaschi, Felix Cheng, Jun Bin Lee, En-Shiun Annie |
| contents | Realignment is a promising strategy to improve cross-lingual transfer in multilingual language models. However, empirical results are mixed and often unreliable, particularly for typologically distant or low-resource languages (LRLs) compared to English. Moreover, word realignment tools often rely on high-quality parallel data, which can be scarce or noisy for many LRLs. In this work, we conduct an extensive empirical study to investigate whether realignment truly benefits from using all available languages, or if strategically selected subsets can offer comparable or even improved cross-lingual transfer, and study the impact on LRLs. Our controlled experiments show that realignment can be particularly effective for LRLs and that using carefully selected, linguistically diverse subsets can match full multilingual alignment, and even outperform it for unseen LRLs. This indicates that effective realignment does not require exhaustive language coverage and can reduce data collection overhead, while remaining both efficient and robust when guided by informed language selection. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_06497 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Rethinking what Matters: Effective and Robust Multilingual Realignment for Low-Resource Languages Nguyen, Quang Phuoc Anugraha, David Gaschi, Felix Cheng, Jun Bin Lee, En-Shiun Annie Computation and Language Artificial Intelligence Realignment is a promising strategy to improve cross-lingual transfer in multilingual language models. However, empirical results are mixed and often unreliable, particularly for typologically distant or low-resource languages (LRLs) compared to English. Moreover, word realignment tools often rely on high-quality parallel data, which can be scarce or noisy for many LRLs. In this work, we conduct an extensive empirical study to investigate whether realignment truly benefits from using all available languages, or if strategically selected subsets can offer comparable or even improved cross-lingual transfer, and study the impact on LRLs. Our controlled experiments show that realignment can be particularly effective for LRLs and that using carefully selected, linguistically diverse subsets can match full multilingual alignment, and even outperform it for unseen LRLs. This indicates that effective realignment does not require exhaustive language coverage and can reduce data collection overhead, while remaining both efficient and robust when guided by informed language selection. |
| title | Rethinking what Matters: Effective and Robust Multilingual Realignment for Low-Resource Languages |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2511.06497 |