長度自適應解碼用於遮罩擴散機器翻譯

摘要

機器翻譯對遮罩擴散語言模型（dLLMs）構成考驗，因為每一個來源詞元都必須被忠實地呈現；而固定畫布解碼必須在去噪前先選定目標長度。現有的遮罩擴散解碼研究主要探討詞元的解遮罩順序，卻對這個直接影響涵蓋度與冗餘的長度決定著墨甚少。我們提出 Entropy-Valley（EV），一種免訓練的長度選擇器，它透過全遮罩前向傳遞的平均預測熵來評分候選目標畫布，並選擇骨幹模型最有把握填滿的畫布。相較於使用訓練語料長度統計的基線，EV 在 EntoZh、ZhtoEn 與 EntoDe 上分別恢復了參考目標長度所帶來之 COMET-22 增益的 64.9%、65.3% 與 33.0%。我們的診斷顯示，有利於去噪的長度不必與參考長度相符。三位翻譯專家進行的評估支持英中（En↔Zh）方向的充分性增益，其中在 ZhtoEn 上的證據更為明顯。與使用相同微調資料訓練的 LLaMA-3-8B 自迴歸（AR）模型相比，EV 系統在 EntoZh 上持平，在 ZhtoEn 上領先；一項以 oracle 長度進行的診斷進一步顯示，在此遮罩擴散機器翻譯設定中，先揭露哪些詞元的決定，其影響不如目標長度的提供方式來得重要。

English

Machine translation tests masked diffusion language models (dLLMs) because every source token must be rendered faithfully, while fixed canvas decoding must choose target length before denoising. Existing masked diffusion decoding work mainly studies token unmasking order, leaving this length decision under-explored despite its direct effect on coverage and redundancy. We introduce Entropy-Valley (EV), a training-free length selector that scores candidate target canvases by mean predictive entropy from all-mask forward passes and selects the canvas the backbone is most prepared to fill. Relative to a baseline using training corpus length statistics, EV recovers 64.9%, 65.3%, and 33.0% of the COMET-22 gain from reference target lengths on EntoZh, ZhtoEn, and EntoDe. Our diagnostics show that denoising-friendly lengths need not match reference lengths. Evaluation by three translation experts supports the EnleftrightarrowZh adequacy gains, with stronger evidence on ZhtoEn. Compared with a LLaMA-3-8B autoregressive (AR) model trained on the same fine-tuning data, the EV system ties on EntoZh and leads on ZhtoEn; an oracle-length diagnostic further shows that, in this masked diffusion MT setting, deciding which tokens to reveal first matters less than how the target length is supplied.