研究人员推出了一种名为边际条件桥的新型流语言模型（FLM）采样方法。该技术将连续流匹配应用于 token 序列，解决了标准扩散模型采样器的局限性。所提出的方法从 FLM token 边际中采样端点，然后使用解析 Ornstein-Uhlenbeck 桥，从而在质量-多样性权衡方面得到改进，并对解码进行原则性控制。 AI
影响 引入了一种新颖的采样技术，增强了流语言模型的质量-多样性平衡。
排序理由 该集群包含一篇学术论文，详细介绍了一种从特定类型语言模型中进行采样的新方法。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
Flow Language Models (FLMs) are a recently introduced class of language models which adapt continuous flow matching for one-hot encoded token sequences. Their denoisers have a special structure absent from generic continuous diffusion models: each block of the denoising mean is a…
arXiv:2605.13681v1 Announce Type: cross Abstract: Flow Language Models (FLMs) are a recently introduced class of language models which adapt continuous flow matching for one-hot encoded token sequences. Their denoisers have a special structure absent from generic continuous diffu…