Document-Level MT

MT: Machine Translation

Document-level machine translation studies how to use context across sentences and paragraphs to resolve references, maintain terminology and style, and preserve document structure. Approaches include chunking long texts, selecting relevant history with retrieval or memory, and adapting large language models through fine-tuning, preference learning, or agent-based methods. Evaluation increasingly considers discourse coherence, sentence alignment, terminology, and layout, alongside conventional translation quality benchmarks. A central tension is how to provide enough context for consistency without adding noise, excessive cost, or losses in local accuracy; sentence-level metrics often miss these trade-offs.

Latest papers

20 papers4 papers in the last four weeks, up 33% on the four weeks before. 0.0% of all new papers.

Retrieve What Helps, Only When It Helps: Agentic Reinforcement Learning for Selective Retrieval-Augmented Machine Translation
Oct 14, 2026Shijun Li, Yang Liu, Najmeh Sadoughi +5Machine Translation Retrieval-Augmented Generation
The University of Texas at Austin · Amazon
Index-Translate: A Multilingual Translation Model Family -- Text, Speech, Controlled Dubbing, and Long-Document Translation
Sep 30, 2026Tianjiao Li, Mengran Yu, Chenyu Shi +8Machine Translation Speech Translation
Index LLM Team
Breaking Babel: A Self-Evolving Multi-Agent System for Long-Form Subtitle Translation
Sep 29, 2026Haibo Jin, Xinjie Li, Najmeh Sadoughi +4Multi-Agent LLM Systems Document-Level MT
University of Illinois Urbana-Champaign, USA · Amazon, USA
TermJudge: A Document-Level Metric Judging, Not Counting, Terminology in Machine Translation Evaluation
Sep 28, 2026Nicolas Dahan, Fran{\cc}ois Yvon, Rachel BawdenAutomated Evaluation Document-Level MT
ISIR, ALMAnaCH · MLIA, ISIR · ALMAnaCH
Doc2FRC: Length-Consistent Document-Level Machine Translation via Fixed-Range Chunking
Sep 11, 2026Xiaotian Wang, Youyuan Lin, Zhan Shen +1Document Chunking LLM Fine-Tuning
The University of Tokyo · Riken · Kyoto University +1
Discourse Dependency: A Continuous Criterion for Translation Difficulty
Sep 4, 2026Ahrii Kim, Chanjun Park, Seong-heum KimMT Evaluation Context-Aware MT
AI-Bio Convergence Research Inst. Soongsil University · School of Software Soongsil University · Dept. of Intelligent Semiconductors Soongsil University
DiscoSign: Discourse-Aware Text to Sign Language Gloss Translation
Sep 2, 2026Vasileios Baltatzis, Mert Inan, Connor Gillis +4Document-Level MT Coreference Resolution
Apple · Northeastern University
STAR : Sentence Translation Alignment Rate for Document-to-Document Machine Translation
Aug 27, 2026Yichen Dong, Hao Wang, Junhui Li +3Machine Translation Document-Level MT
School of Computer Science and Technology, Soochow University, Suzhou, China · Alibaba Group, Hangzhou, China
APEX-VW: A Document-Level English-Spanish Post-Editing Dataset in the Healthcare Domain
Aug 8, 2026Marie Escribe, Tharindu Ranasinghe, Amal Haddad Haddad +2Document-Level MT Machine Translation
Universitat Politècnica de València · València, Spain · Lancaster University +3
Studying quantization trade-offs for efficient inference deployment in machine translation
Jul 31, 2026Jim Zhao, Sohir Maskey, Koen Oostermeijer +2LLM Quantization LLM Inference Acceleration
University of Basel · Aleph Alpha Research · Graphcore
Can an Old Dog Be Taught New Tricks? Taking LLMs Beyond Sentence Level Translation
Jul 15, 2026Alaina BrandtLLM Prompting Document-Level MT
Structure-Preserving Document Translation via Multi-Stage LLM Pipeline: A Case Study in Marathi
Jun 27, 2026Manasi Waghe, Danish Chandargi, Mohammad Aamir Rayyan +2Document-Level MT Machine Translation
Pune Institute of Computer Technology, Pune, India · L3Cube Labs, Pune, India · Indian Institute of Technology Madras, Chennai, India
CzechDocs: A Multiway Parallel Dataset of Formatted Documents for Minority Languages in Czechia
Jun 18, 2026Josef Jon, Ondřej BojarDocument-Level MT Machine Translation
Charles University, Faculty of Mathematics and Physics Institute of Formal and Applied Linguistics Prague, Czech Republic
G^2C-MT: Graph-Guided Context Selection for Document-Level Machine Translation
Jun 2, 2026Baijun Ji, Zixuan Zhou, Xiangyu Duan +4Graph-Based Retrieval Document-Level MT
School of Computer Science and Technology, Soochow University
Loong: A Human-Like Long Document Translation Agent with Observe-and-Act Adaptive Context Selection
May 28, 2026Yutong Wang, Xuebo Liu, Derek F. Wong +6Adaptive Retrieval Document-Level MT
Institute of Computing and Intelligence, Harbin Institute of Technology, Shenzhen · 2NLP2CT Lab, Department of Computer and Information Science, University of Macau · 3Huawei Translation Services Center
What Does LLM Refinement Actually Improve? A Systematic Study on Document-Level Literary Translation
May 13, 2026Shaomu Tan, Dawei Zhu, Ke Tran +5LLM Self-Refinement Document-Level MT
University of Amsterdam · 3Amazon AGI · University of Cambridge
BabelDOC: Better Layout-Preserving PDF Translation via Intermediate Representation
May 11, 2026Qi Yang, Xiangyao Ma, Xiao Wang +2Document-Level MT Machine Translation
School of Computer Engineering and Science, Shanghai University, Shanghai, China · Funstory.ai Limited, Hong Kong SAR, China · Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai, China
Cross-Preference Learning for Sentence-Level and Context-Aware Machine Translation
Mar 26, 2026Ying Li, Xinglin Lyu, Junhui Li +5Context-Aware MT Machine Translation
School of Computer Science and Technology, Soochow University, Suzhou, China · School of Computer Science and Artificial Intelligence, Zhengzhou University, Zhengzhou, China · Jiangsu Key Lab of Language Computing, China +1
Enhancing Document-Level Machine Translation via Filtered Synthetic Corpora and Two-Stage LLM Adaptation
Mar 23, 2026Ireh Kim, Tesia Sker, Chanwoo KimMachine Translation Document-Level MT
Department of Artificial Intelligence, Korea University, Seoul, South Korea
ICDAR 2025 Competition on End-to-End Document Image Machine Translation Towards Complex Layouts
Mar 10, 2026Yaping Zhang, Yupu Liang, Zhiyang Zhang +5Image Text Translation Machine Translation
Institute of Automation, Chinese Academy of Sciences · University of Chinese Academy of Sciences, Beijing, China · Fanyu AI Laboratory, Zhongke Fanyu Technology Company Ltd., Beijing, China.

Oct 14, 2026J/K move · Enter open · S save

Retrieve What Helps, Only When It Helps: Agentic Reinforcement Learning for Selective Retrieval-Augmented Machine Translation

Shijun Li, Yang Liu, Najmeh Sadoughi, Dongmei Wang+4

The University of Texas at Austin · Amazon

Machine Translation Retrieval-Augmented Generation Adaptive RAG RL for IR Document-Level MT

Open paper AskPDF arXiv

Retrieving external knowledge can improve large language model (LLM) translation by providing supplementary information such as idiom interpretations and terminology. However, existing retrieval-augmented machine translation systems typically apply predefined retrieval strategies uniformly across all inputs. We argue that translation operates in a partial-evidence regime: LLM's parametric knowledge suffices for many inputs, while external evidence benefits a subset and usually supports rather than directly provides the desired translation. This creates two challenges: deciding when to retrieve and what evidence to acquire. We introduce Selective Retrieval-Augmented Translation (SRAT), an agentic reinforcement learning framework that learns when to translate directly or retrieve, what to retrieve, and how to integrate these decisions by optimizing final translation quality. To align evidence retrieval with downstream translation quality rather than simple semantic similarity, we further develop joint Plackett--Luce optimization and translation-quality contrastive training for retriever-training. Experiments on idiom, knowledge-intensive, and document-level translation with offline or online retrieval show that SRAT consistently outperforms never- and always-retrieve baselines, and quality-aligned retriever training provides further gains when similarity-based retrieval is insufficient. The results demonstrate the benefit of treating retrieval as a learnable translation action rather than fixed preprocessing.