Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.25836 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914425840074752 |
|---|---|
| author | Sun, Ruiyan Nakamura, Satoshi |
| author_facet | Sun, Ruiyan Nakamura, Satoshi |
| contents | In low-resource multilingual speech-to-text translation, uniform architectural sharing across languages frequently introduces representation conflicts that impede convergence. This work proposes a principled methodology to automatically determine layer-specific sharing patterns by mining training gradient information. Our approach employs three distinct analysis strategies: distance-based language clustering, self/cross-task divergence metrics for capacity allocation, and joint factorization coupled with canonical correlation analysis for subspace alignment. Extensive evaluation across four language pairs (using the SeamlessM4T-Medium architecture) demonstrates persistent improvements in translation quality metrics. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_25836 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Gradient-Informed Training for Low-Resource Multilingual Speech Translation Sun, Ruiyan Nakamura, Satoshi Computation and Language In low-resource multilingual speech-to-text translation, uniform architectural sharing across languages frequently introduces representation conflicts that impede convergence. This work proposes a principled methodology to automatically determine layer-specific sharing patterns by mining training gradient information. Our approach employs three distinct analysis strategies: distance-based language clustering, self/cross-task divergence metrics for capacity allocation, and joint factorization coupled with canonical correlation analysis for subspace alignment. Extensive evaluation across four language pairs (using the SeamlessM4T-Medium architecture) demonstrates persistent improvements in translation quality metrics. |
| title | Gradient-Informed Training for Low-Resource Multilingual Speech Translation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2603.25836 |