Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.25183 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918410662707200 |
|---|---|
| author | Li, Ying Lyu, Xinglin Li, Junhui Yang, Jinlong Shang, Hengchao Zhang, Min Tao, Shimin Wei, Daimeng |
| author_facet | Li, Ying Lyu, Xinglin Li, Junhui Yang, Jinlong Shang, Hengchao Zhang, Min Tao, Shimin Wei, Daimeng |
| contents | Context-aware machine translation (MT) leverages document-level information, yet it does not consistently outperform sentence-level MT, as contextual signals are unevenly beneficial across sentences. Existing training objectives do not explicitly model this variability, limiting a model's ability to adaptively exploit context. In this paper, we propose Cross-Preference Learning (CPL), a preference-based training framework that explicitly captures the complementary benefits of sentence-level and context-aware MT. CPL achieves this by integrating both intra- and cross-condition preferences into the preference optimization objective. The introduction of intra- and cross-condition preferences provides explicit supervision on when and how contextual information improves translation quality. We validate the proposed approach on several public context-aware MT tasks using multiple models, including Qwen3-4B, Qwen3-8B, and Llama-3-8B. Experimental results demonstrate consistent improvements in translation quality and robustness across both input conditions, achieved without any architectural modifications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_25183 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Cross-Preference Learning for Sentence-Level and Context-Aware Machine Translation Li, Ying Lyu, Xinglin Li, Junhui Yang, Jinlong Shang, Hengchao Zhang, Min Tao, Shimin Wei, Daimeng Computation and Language Context-aware machine translation (MT) leverages document-level information, yet it does not consistently outperform sentence-level MT, as contextual signals are unevenly beneficial across sentences. Existing training objectives do not explicitly model this variability, limiting a model's ability to adaptively exploit context. In this paper, we propose Cross-Preference Learning (CPL), a preference-based training framework that explicitly captures the complementary benefits of sentence-level and context-aware MT. CPL achieves this by integrating both intra- and cross-condition preferences into the preference optimization objective. The introduction of intra- and cross-condition preferences provides explicit supervision on when and how contextual information improves translation quality. We validate the proposed approach on several public context-aware MT tasks using multiple models, including Qwen3-4B, Qwen3-8B, and Llama-3-8B. Experimental results demonstrate consistent improvements in translation quality and robustness across both input conditions, achieved without any architectural modifications. |
| title | Cross-Preference Learning for Sentence-Level and Context-Aware Machine Translation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2603.25183 |