Safety
STAR : Sentence Translation Alignment Rate for Document-to-Document Machine Translation
arXiv:2608.27161v1 Announce Type: new Abstract: Large Language Models (LLMs) have enabled a shift from sentence-level to document-to-document (Doc2Doc) machine translation, promising improved global c
arXiv:2608.27161v1 Announce Type: new Abstract: Large Language Models (LLMs) have enabled a shift from sentence-level to document-to-document (Doc2Doc) machine translation, promising improved global coherence. However, document-to-document generation in a single pass frequently suffers from structural misalignment, manifesting as sentence omissions or hallucinations that violate the core requirement of source-target correspondence. To address this, we introduce Sentence Translation Alignment Rate (STAR), an auxiliary metric that explicitly quantifies sentence-level structural fidelity. Building on this, we propose STAR-masked Preference Optimization (StarPO), a framework that ranks document-level hypotheses by structural quality and utilizes a dynamic alignment mask to focus optimization on misaligned segments. Experimental results across news and literary domains demonstrate that StarPO significantly enhances translation quality and structural integrity. Notably, StarPO allows compact models to surpass the performance of massive proprietary systems like GPT-4o while maintaining superior token efficiency.
Related
- Predicting Multilingual Classification and Translation Performance of LLMs with Cross-Lingual Alignment nicode{x2013} Is English Enough?
- Distributional Alignment as a Criterion for Designing Task Vectors in In-Context Learning
- Language on Demand, Knowledge at Core: Composing LLMs with Encoder-Decoder Translation Models for Extensible Multilinguality
Source: arXiv cs.CL | 2026-08-28