发表机构
Peking University; BYD Company Limited（北京大学; 比亚迪股份有限公司）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对掩码扩散机器翻译的长度决策问题，提出无训练的Entropy-Valley长度选择器，在多个翻译方向上显著优于基线，性能接近同数据训练的LLaMA-3-8B自回归模型。
AI 中文摘要
机器翻译对掩码扩散语言模型（dLLM）构成考验，因为每个源词元都必须被忠实地呈现，而固定画布解码需要在去噪前选定目标长度。现有掩码扩散解码研究主要关注词元的去掩码顺序，却未充分探索长度决策，尽管它对覆盖率和冗余度有直接影响。我们提出了无训练的长度选择器Entropy-Valley（EV），该方法通过全掩码前向传播得到的平均预测熵对候选目标画布进行评分，选择主干模型最有能力填充的画布。与使用训练语料长度统计的基线相比，EV在英→中、中→英、英→德方向上分别恢复了参考目标长度所带来的COMET-22增益的64.9%、65.3%和33.0%。我们的分析表明，适合去噪的长度不必与参考长度匹配。三名翻译专家的评估支持英↔中方向的充分性提升，其中中→英方向的证据更充分。与在相同微调数据上训练的LLaMA-3-8B自回归（AR）模型相比，EV系统在英→中方向表现相当，在中→英方向领先；最优长度分析进一步显示，在该掩码扩散机器翻译场景中，先揭示哪些词元的重要性低于如何提供目标长度。
英文摘要
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 En$\to$Zh, Zh$\to$En, and En$\to$De. Our diagnostics show that denoising-friendly lengths need not match reference lengths. Evaluation by three translation experts supports the En$\leftrightarrow$Zh adequacy gains, with stronger evidence on Zh$\to$En. Compared with a LLaMA-3-8B autoregressive (AR) model trained on the same fine-tuning data, the EV system ties on En$\to$Zh and leads on Zh$\to$En; 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.
CommentsAccepted to the Main Conference of EMNLP 2026. 22 pages, 7 figures. Code: https://github.com/Entropy-Valley/Entropy-Valley ; dataset and model: https://huggingface.co/collections/YanZhanPKU/entropy-valley
