English(EN) DELA: Learning Controllable Edits for Diffusion Language Models

DELA框架实现了扩散语言模型的可控编辑

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-26 04:00

研究人员推出DELA，一个旨在增强扩散语言模型（DLMs）可控编辑能力的新框架。DELA作为一个插件运行，通过插入、删除和重新遮蔽等操作实现显式的草稿精炼，同时保持核心DLM冻结。该方法将高级编辑控制与令牌级生成分离，通过采用两阶段解码过程来解决计算效率低下问题，该过程在草稿阶段优化KV缓存重用，并在后续精炼阶段允许有针对性的非单调编辑。实验表明，DELA可以将数学推理和代码生成等任务的性能提高多达5.5个百分点。 AI

影响增强了扩散模型中文本生成的控制能力，可能提高代码生成和数学推理等任务的性能。

排序理由该集群包含一篇详细介绍扩散语言模型新框架的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 12 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇详细介绍扩散语言模型新框架的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Xinsong Feng, Chenan Wang, Leshu Tang, Johnny Liu, Dong Li, Da Li, Haipeng Chen · 2026-10-26 04:00

DELA：为扩散语言模型学习可控编辑

arXiv:2610.32096v1 Announce Type: new Abstract: Diffusion language models (DLMs) support bidirectional denoising and iterative refinement, but generation remains largely controlled by token-level masking schedules. This leaves draft-level refinement implicit and limits explicit c…

本文如何被排名

报道来源 [1]

DELA：为扩散语言模型学习可控编辑

相关实体

相关话题