mlsanityData to 7 Oct 2026

MERIT: Multilingual Expert-Reward Informed Tuning for Chinese-Centric Low-Resource Machine Translation

arXiv 2604.04839 ↗First posted 6 Apr 2026cs.CLcited by 1

Not on the map. Why?

Abstract (arXiv)

Neural machine translation (NMT) from Chinese to low-resource Southeast Asian languages remains severely constrained by the extreme scarcity of clean parallel corpora and the pervasive noise in existing mined data. This chronic shortage not only impedes effective model training but also sustains a large performance gap with high-resource directions, leaving millions of speakers of languages such as Lao, Burmese, and Tagalog with persistently low-quality translation systems despite recent advances in large multilingual models. We introduce Multilingual Expert-Reward Informed Tuning (MERIT), a unified translation framework that transforms the traditional English-centric ALT benchmark into a Chinese-centric evaluation suite for five Southeast Asian low-resource languages (LRLs). Our framework combines language-specific token prefixing (LTP) with supervised fine-tuning (SFT) and a novel group relative policy optimization (GRPO) guided by the semantic alignment reward (SAR). These results confirm that, in LRLtextrightarrowChinese translation, targeted data curation and reward-guided optimization dramatically outperform mere model scaling.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 33 of its 65 references that are in the map. First posted on arXiv 6 Apr 2026.

Builds on

33 of 65 references.

References · 33

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
  3. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
  4. 2022OPT: Open Pre-trained Transformer Language Modelscited by 3,464
  5. 2014Sequence to Sequence Learning with Neural Networkscited by 2,974
  6. 2020mT5: A massively multilingual pre-trained text-to-text transformercited by 1,813
  7. 2018A Call for Clarity in Reporting BLEU Scorescited by 1,790
  8. 2018Datasheets for Datasetscited by 1,371
  9. 2018Model Cards for Model Reportingcited by 1,068
  10. 2020Multilingual Denoising Pre-training for Neural Machine Translationcited by 1,028
  11. 2020COMET: A Neural Framework for MT Evaluationcited by 1,024
  12. 2020Beyond English-Centric Multilingual Machine Translationcited by 559
21 more
  1. 2021The FLORES-101 Evaluation Benchmark for Low-Resource and Multilingual Machine Translationcited by 556
  2. 2016Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translationcited by 467
  3. 2018Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyondcited by 434
  4. 2021Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translationcited by 333
  5. 2019Massively Multilingual Neural Machine Translation in the Wild: Findings and Challengescited by 253
  6. 2024From LLM to NMT: Advancing Low-Resource Machine Translation with Claudecited by 57
  7. 2021DeltaLM: Encoder-Decoder Pre-training for Language Generation and Translation by Augmenting Pretrained Multilingual Encoderscited by 50
  8. 2018Dual Conditional Cross-Entropy Filtering of Noisy Parallel Corporacited by 40
  9. 2025Multilingual Machine Translation with Open Large Language Models at Practical Scale: An Empirical Studycited by 38
  10. 2024TEaR: Improving LLM-based Machine Translation with Systematic Self-Refinementcited by 37
  11. 2024TransLLaMa: LLM-based Simultaneous Translation Systemcited by 34
  12. 2024Mitigating the Language Mismatch and Repetition Issues in LLM-based Machine Translation via Model Editingcited by 18
  13. 2024Aligning Large Language Models via Fine-grained Supervisioncited by 13
  14. 2025Compositional Translation: A Novel LLM-based Approach for Low-resource Machine Translationcited by 13
  15. 2024Ladder: A Model-Agnostic Framework Boosting LLM-based Machine Translation to the Next Levelcited by 11
  16. 2020scb-mt-en-th-2020: A Large English-Thai Parallel Corpuscited by 10
  17. 2024Efficient Terminology Integration for LLM-based Translation in Specialized Domainscited by 7
  18. 2024PromptRefine: Enhancing Few-Shot Performance on Low-Resource Indic Languages with Example Selection from Related Example Bankscited by 5
  19. 2025DeepSeek-R1 vs. o3-mini: How Well can Reasoning LLMs Evaluate MT and Summarization?cited by 5
  20. 2026SAGE: Sustainable Agent-Guided Expert-tuning for Culturally Attuned Translation in Low-Resource Southeast Asiacited by 2
  21. 2026EcoThink: A Green Adaptive Inference Framework for Sustainable and Accessible Agentscited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Semantic-Topological Graph Reasoning for Language-Guided Pulmonary Screeningcited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2025FuxiMT: Sparsifying Large Language Models for Chinese-Centric Multilingual Machine Translation
  2. 2026Reinforcement Learning with Semantic Rewards Enables Low-Resource Language Expansion without Alignment Tax
  3. 2026Reference-Free Reinforcement Learning Fine-Tuning for MT: A Seq2Seq Perspective
  4. 2025SSR-Zero: Simple Self-Rewarding Reinforcement Learning for Machine Translation
  5. 2026Reference-Free Post-Training of Open Large Language Models for Multilingual Machine Translation
  6. 2024A Preference-driven Paradigm for Enhanced Translation with Large Language Models
  7. 2025Improving Low-Resource Translation with Dictionary-Guided Fine-Tuning and RL: A Spanish-to-Wayuunaiki Study
  8. 2025Bridging the Linguistic Divide: A Survey on Leveraging Large Language Models for Machine Translation