ALoDLM：自适应循环扩散语言模型
原标题：Paper page - ALoDLM: Adaptively Looped Diffusion Language Models
AI 摘要
该论文提出 ALoDLM，一种自适应循环扩散语言模型，用 token 级自适应潜在循环替代均匀计算深度，根据 token 难度分配计算量。作者将 token 级计算调度建模为隐变量并推导条件 NELBO，联合学习 token 预测与计算分配。在 1.7B 和 8B 参数规模、十一个基准上，ALoDLM 平均得分超过所有评估的扩散模型及对应自回归基线，同时保持快速并行解码，取得较好的质量-效率权衡。
正文节选
ALoDLM: Adaptively Looped Diffusion Language Models Abstract Diffusion language models (DLMs) enable fast generation by predicting multiple tokens in parallel, but their practical adoption remains limited by a persistent quality gap relative to comparably sized autoregressive (AR) models. We attribute this gap to a computation-difficulty mismatch: within a partially observed sequence, some unknown tokens are easy to predict, while others require substantially more computation. Existing DLMs neve