Pith. sign in

REVIEW 8 cited by

MT-R1-Zero: Advancing LLM-based Machine Translation via R1-Zero-like Reinforcement Learning

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2504.10160 v1 pith:RM3IPZ72 submitted 2025-04-14 cs.CL cs.AIcs.LG

Zhaopeng Feng , Shaosheng Cao , Jiahan Ren , Jiayuan Su , Ruizhe Chen , Yan Zhang , Zhe Xu , Yao Hu
show 2 more authors
Jian Wu Zuozhu Liu
This is my paper
classification cs.CLcs.AIcs.LG
keywords metricsmt-r1-zeroreasoningrewardtranslationachievesacrossaverage
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Large-scale reinforcement learning (RL) methods have proven highly effective in enhancing the reasoning abilities of large language models (LLMs), particularly for tasks with verifiable solutions such as mathematics and coding. However, applying this idea to machine translation (MT), where outputs are flexibly formatted and difficult to automatically evaluate with explicit rules, remains underexplored. In this work, we introduce MT-R1-Zero, the first open-source adaptation of the R1-Zero RL framework for MT without supervised fine-tuning or cold-start. We propose a rule-metric mixed reward mechanism to guide LLMs towards improved translation quality via emergent reasoning. On the WMT 24 English-Chinese benchmark, our MT-R1-Zero-3B-Mix achieves competitive performance, surpassing TowerInstruct-7B-v0.2 by an average of 1.26 points. Meanwhile, our MT-R1-Zero-7B-Mix attains a high average score of 62.25 across all metrics, placing it on par with advanced proprietary models such as GPT-4o and Claude-3.5-Sonnet, while the MT-R1-Zero-7B-Sem variant achieves state-of-the-art scores on semantic metrics. Moreover, our work exhibits strong generalization capabilities on out-of-distribution MT tasks, robustly supporting multilingual and low-resource settings. Extensive analysis of model behavior across different initializations and reward metrics offers pioneering insight into the critical role of reward design, LLM adaptability, training dynamics, and emergent reasoning patterns within the R1-Zero paradigm for MT. Our code is available at https://github.com/fzp0424/MT-R1-Zero.

Discussion (0). Sign in to comment.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Seed LiveInterpret 2.0: End-to-end Simultaneous Speech-to-speech Translation with Your Voice

    cs.CL 2025-07 conditional novelty 6.0 of 10

    An end-to-end simultaneous speech-to-speech translation model with voice cloning, trained with a two-stage reinforcement learning reward scheme, reports high accuracy and low latency on the authors' RealSI benchmark.

  2. TAT-R1: Terminology-Aware Translation with Reinforcement Learning and Word Alignment

    cs.CL 2025-05 conditional novelty 6.0 of 10

    Word-alignment rewards for RL-trained translation raise terminology accuracy on RTT from 54.42 to 56.42 TA without hurting general translation quality.

  3. MT$^{3}$: Scaling MLLM-based Text Image Machine Translation via Multi-Task Reinforcement Learning

    cs.CL 2025-05 conditional novelty 6.0 of 10

    A 7B multimodal model trained with multi-task reinforcement learning beats much larger models on image-text translation benchmarks, though some out-of-distribution claims are contradicted by the paper's own tables.

  4. Reasoning Before Translation: Enhancing Legal Machine Translation with Structured Reasoning

    cs.CL 2026-07 conditional novelty 5.0 of 10

    On Swiss legal translation, reinforcement learning with a ChrF reward improves small open models more than supervised fine-tuning, but frontier reasoning models still score higher.

  5. AutoTIR: Autonomous Tools Integrated Reasoning via Reinforcement Learning

    cs.CL 2025-07 conditional novelty 5.0 of 10

    AutoTIR applies GRPO with a hand-designed action reward so a 7B instruct model learns to mix search and code tools, beating tool-using baselines on ten benchmarks.

  6. Datasets and Recipes for Video Temporal Grounding via Reinforcement Learning

    cs.CV 2025-07 conditional novelty 5.0 of 10

    A supervised fine-tuning plus difficulty-filtered reinforcement learning recipe improves video temporal grounding on three benchmarks, with datasets and models released.

  7. RIVAL: Reinforcement Learning with Iterative and Adversarial Optimization for Machine Translation

    cs.CL 2025-06 conditional novelty 5.0 of 10

    RIVAL iteratively re-trains a reward model adversarially against the current translator and adds a BLEU-predicting head, improving in-domain WMT and subtitle translation over SFT baselines.

  8. TACTIC: Translation Agents with Cognitive-Theoretic Interactive Collaboration

    cs.CL 2025-06 conditional novelty 4.0 of 10

    TACTIC, a cognitive-inspired six-agent workflow, improves LLM translation quality over direct prompting on FLORES-200 and WMT24, with the best DeepSeek-V3 setup reaching 96.19 XCOMET on English-to-X.

Pith tools