用蒙特卡洛树搜索优化扩散模型的填槽顺序，提升生成质量。
Can I Have Your Order? Monte-Carlo Tree Search for Slot Filling Ordering in Diffusion Language Models
- 将填槽顺序建模为决策问题，用MCTS探索最优生成路径。
- 在MBPP上提升19.5%，平均比自回归方法高3.2%。
- 适合需要高质量推理生成的研究者和开发者。
尽管掩码扩散模型（MDMs）在数学与代码推理中展现潜力，其性能仍高度依赖填槽顺序，导致输出方差大。我们提出McDiffuSE框架，将槽选择建模为决策问题，并通过蒙特卡洛树搜索（MCTS）优化填槽顺序。McDiffuSE利用前瞻模拟评估部分完成结果，系统探索生成顺序的组合空间。实验显示，相比自回归基线平均提升3.2%，相比基线计划-填充方法提升8.0%，在MBPP上达19.5%、MATH500上达4.9%。分析表明，虽多数情况下采用顺序生成，但引入非顺序生成对性能提升至关重要。更大的探索常数而非更多模拟，是克服模型置信度偏差、发现有效顺序的关键。这些发现确立了基于MCTS的规划在提升MDMs生成质量上的有效性。
原文摘要 · Abstract (English)
While plan-and-infill decoding in Masked Diffusion Models (MDMs) shows promise for mathematical and code reasoning, performance remains highly sensitive to slot infilling order, often yielding substantial output variance. We introduce McDiffuSE, a framework that formulates slot selection as decision making and optimises infilling orders through Monte Carlo Tree Search (MCTS). McDiffuSE uses look-ahead simulations to evaluate partial completions before commitment, systematically exploring the combinatorial space of generation orders. Experiments show an average improvement of 3.2% over autoregressive baselines and 8.0% over baseline plan-and-infill, with notable gains of 19.5% on MBPP and 4.9% on MATH500. Our analysis reveals that while McDiffuSE predominantly follows sequential ordering, incorporating non-sequential generation is essential for maximising performance. We observe that larger exploration constants, rather than increased simulations, are necessary to overcome model confidence biases and discover effective orderings. These findings establish MCTS-based planning as an effective approach for enhancing generation quality in MDMs.
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。