mlsanityData to 14 Oct 2026

Trans-Zero: Self-Play Incentivizes Large Language Models for Multilingual Translation Without Parallel Data

arXiv 2504.14669 ↗First posted 20 Apr 2025cs.CLcited by 4

On the map: Synthetic data and self-improvement for LLMs → LLM synthetic data generation

Abstract (arXiv)

The rise of Large Language Models (LLMs) has reshaped machine translation (MT), but multilingual MT still relies heavily on parallel data for supervised fine-tuning (SFT), facing challenges like data scarcity for low-resource languages and catastrophic forgetting. To address these issues, we propose TRANS-ZERO, a self-play framework that leverages only monolingual data and the intrinsic multilingual knowledge of LLM. TRANS-ZERO combines Genetic Monte-Carlo Tree Search (G-MCTS) with preference optimization, achieving strong translation performance that rivals supervised methods. Experiments demonstrate that this approach not only matches the performance of models trained on large-scale parallel data but also excels in non-English translation directions. Further analysis reveals that G-MCTS itself significantly enhances translation quality by exploring semantically consistent candidates through iterative translations, providing a robust foundation for the framework's succuss.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Placed in LLM synthetic data generation, one of 13 directions in Synthetic data and self-improvement for LLMs. Builds on 19 of its 26 references that are in the map. First posted on arXiv 20 Apr 2025.

Builds on

19 of 26 references.

References · 19

  1. 2017Attention Is All You Needcited by 60,361
  2. 2023GPT-4 Technical Reportcited by 18,396
  3. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,780
  4. 2022No Language Left Behind: Scaling Human-Centered Machine Translationcited by 1,312
  5. 2020BLEURT: Learning Robust Metrics for Text Generationcited by 1,203
  6. 2022Large Language Models Can Self-Improvecited by 664
  7. 2024Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Modelscited by 557
  8. 2024Training Language Models to Self-Correct via Reinforcement Learningcited by 426
  9. 2024Contrastive Preference Optimization: Pushing the Boundaries of LLM Performance in Machine Translationcited by 394
  10. 2024ReST-MCTS*: LLM Self-Training via Process Reward Guided Tree Searchcited by 344
  11. 2022CometKiwi: IST-Unbabel 2022 Submission for the Quality Estimation Shared Taskcited by 207
  12. 2024Tower: An Open Multilingual Large Language Model for Translation-Related Taskscited by 206
7 more
  1. 2021Monte Carlo Tree Search: A Review of Recent Modifications and Applicationscited by 161
  2. 2024Toward Self-Improvement of LLMs via Imagination, Searching, and Criticizingcited by 125
  3. 2024MAPO: Advancing Multilingual Reasoning through Multilingual Alignment-as-Preference Optimizationcited by 69
  4. 2024X-ALMA: Plug & Play Modules and Adaptive Rejection for Quality Translation at Scalecited by 36
  5. 2024MT-PATCHER: Selective and Extendable Knowledge Distillation from Large Language Models for Machine Translationcited by 10
  6. 2024Enhancing Translation Accuracy of Large Language Models through Continual Pre-Training on Parallel Datacited by 7
  7. 2024Why Not Transform Chat Large Language Models to Non-English?cited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 1.

4 papers.

  1. 2025Seed-X: Building Strong Multilingual Translation LLM with 7B Parameterscited by 21
  2. 2025SSR-Zero: Simple Self-Rewarding Reinforcement Learning for Machine Translationcited by 13
  3. 2025DuPO: Enabling Reliable LLM Self-Verification via Dual Preference Optimizationcited by 7
  4. 2026When LLM Meets Tree Search: A Systematic View of Inference as Search in Large Language Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024A Preference-driven Paradigm for Enhanced Translation with Large Language Models
  2. 2024GenTranslate: Large Language Models are Generative Multilingual Speech and Machine Translators
  3. 2025Bridging the Linguistic Divide: A Survey on Leveraging Large Language Models for Machine Translation
  4. 2025The Art of Asking: Multilingual Prompt Optimization for Synthetic Data
  5. 2025MT-R1-Zero: Advancing LLM-based Machine Translation via R1-Zero-like Reinforcement Learning
  6. 2026CAT-Translate: Building Compact Open-Source Models for Japanese-English Translation
  7. 2026CLewR: Curriculum Learning with Restarts for Machine Translation Preference Learning
  8. 2023Do Multilingual Language Models Think Better in English?