提出D5P4方法，提升扩散模型文本生成的多样性与覆盖率。
D5P4: Partition Determinantal Point Process for Diversity in Parallel Discrete Diffusion Decoding
- 用分块确定性点过程建模中间序列选择，实现质量与多样性的内生平衡。
- 在开放生成、问答和数学推理任务上，提升pass@k覆盖率，保持生成质量。
- 无需外部评估器，适用于追求多样性和覆盖的文本生成场景。
离散扩散模型是文本生成的有前景替代方案，但其解码方法仍研究不足。标准自回归搜索（如束搜索）不适用于迭代去噪过程，因为候选序列是完整的中间结果而非从左到右的前缀。此外，现有扩散解码方法对保留序列的多样性和覆盖范围控制有限。本文提出D5P4，一种面向离散扩散模型的束搜索式解码方法，将中间束选择建模为分块确定性点过程下的最大后验推断。该方法生成模型内部的批量目标函数，无需外部验证器即可平衡质量与多样性。在开放式生成、问答和数学推理任务上的实验表明，D5P4在保持或超越基线质量的同时，显著提升了多样性和pass@$k$覆盖率。
原文摘要 · Abstract (English)
Discrete diffusion models are promising alternatives to autoregressive approaches for text generation, yet their decoding methods remain under-studied. Standard autoregressive search procedures, such as beam search, do not directly apply to iterative denoising, where hypotheses are complete intermediate sequences rather than left-to-right prefixes. Furthermore, existing diffusion decoding procedures only provide limited control over the diversity and coverage of retained hypotheses. In this work, we introduce D5P4, a beam-style decoding method tailored to discrete diffusion models, which casts intermediate beam selection as MAP inference under a partitioned Determinantal Point Process. This yields a model-internal batch objective that balances quality and diversity without external verifiers. Experiments on open-ended generation, question answering, and mathematical reasoning show that D5P4 improves diversity and pass@$k$ coverage while matching or surpassing baseline quality and fidelity
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。