@inproceedings{liu-etal-2026-learning-translate,
    title = "Learning to Translate by Translating: Stabilizing the Dual Loop via Semantic-Aware Self-Evolution",
    author = "Liu, Kui  and
      Yin, Mingming  and
      Tang, Zuoli  and
      Li, Zihao  and
      Fu, Chilin  and
      Zhang, Xiaolu  and
      Zhou, Jun  and
      Zou, Lixin  and
      Li, Chenliang",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.findings-acl.723/",
    pages = "14711--14726",
    ISBN = "979-8-89176-395-1",
    abstract = "Despite the remarkable success of Large Language Models (LLMs) in Machine Translation (MT), the scarcity of high-quality parallel corpora and the prohibitive cost of their acquisition constrain scalability. To this end, we propose \textbf{L}earning to \textbf{T}ranslate by \textbf{T}ranslating (\textbf{LTT}), an LLM-driven dual-learning framework that enables autonomous translation, achieving an 80.42{\%} performance improvement over the base model. By adapting the cycle-consistency principle to the generative paradigm, LTT eliminates the need for parallel data. It employs a robust semantic-aware reward function that balances adequacy with reconstruction fidelity, effectively mitigating the reward hacking issues inherent in traditional unsupervised MT. Relying solely on monolingual data, our 8B model consistently outperforms significantly larger models (70B+) in low-resource settings and achieves parity with state-of-the-art supervised baselines on mainstream benchmarks. LTT thus offers a scalable, data-efficient paradigm for autonomous machine translation."
}