提出可直接修正已暴露词元的无参数采样器，提升掩码扩散语言模型性能。
Revise, Don't Freeze: Sampler-Matched Training for Self-Correcting Masked Diffusion Language Models
- 设计无参数采样器D3IM，实现可见词元间的直接修正。
- 在64步去噪下，数学与代码任务准确率提升超10个百分点。
- 适合追求高精度生成、关注推理能力优化的研究者使用。
掩码扩散语言模型（MDLMs）在每一步去噪中重预测所有位置，但标准采样器一旦揭示词元便固定其值，使重修能力被浪费。现有方法要么添加启发式或学习机制修正已确定词元，要么将其重新标记为[MASK]再预测；而能直接修正可见词元且无需额外模块的原理性采样器仍待探索。本文提出D3IM，一种无参数的采样器，基于校正风格的逆向更新，允许直接进行可见到可见的修正，无需额外模块或辅助步骤。D3IM还揭示了模型侧障碍——保留偏差：模型倾向于重复自身错误的已提交词元而非纠正。为此提出SCOPE（Self-Conditioned On Prediction Errors），一种轻量级后训练方法，模拟D3IM的采样过程。在LLaDA-8B模型上，使用64步去噪时，SCOPE+D3IM相较原版标准解码，在GSM8K上提升13.0（达到68.3%），在MATH-500上提升4.8（23.6%），在HumanEval上提升15.3（29.3%），在MBPP上提升10.4（30.8%），且数学与HumanEval任务的增益随去噪步数增加而扩大。
原文摘要 · Abstract (English)
Masked diffusion language models (MDLMs) re-predict every position at each denoising step, but standard samplers commit tokens once revealed, leaving this revision capability unused. Existing approaches either add heuristic or learned mechanisms to revise committed tokens, or remask them back to [MASK] before re-predicting; a principled sampler that directly revises visible tokens without auxiliary modules remains underexplored. We introduce D3IM, a parameter-free sampler derived as a corrector-style reverse update that permits direct visible-to-visible revision without additional modules or auxiliary passes. D3IM also reveals a model-side obstacle we term preservation bias: the model tends to reproduce its own wrong committed tokens rather than correct them. We address this with SCOPE (Self-Conditioned On Prediction Errors), a lightweight post-training procedure that simulates D3IM's sampling process. On LLaDA-8B at 64 denoising steps, SCOPE+D3IM improves over the original LLaDA-8B with standard unmasking by +13.0 on GSM8K (68.3%), +4.8 on MATH-500 (23.6%), +15.3 on HumanEval (29.3%), and +10.4 on MBPP (30.8%), with gains that increase as more denoising steps are used on math and HumanEval.
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。