STARFlow2：连接语言模型与归一化流以实现统一的多模态生成
STARFlow2: Bridging Language Models and Normalizing Flows for Unified Multimodal Generation
AI总结：
本文提出STARFlow2，通过归一化流与预训练视觉语言模型结合，实现高效的多模态生成，采用深度浅层流设计和统一FAE潜在空间，提升生成效率与性能。
摘要：
深度生成模型在文本和视觉领域迅速发展，推动了能够理解、推理和生成交织文本-图像序列的统一多模态系统。现有方法将自回归语言模型与基于扩散的图像生成器结合，继承了因果文本生成与迭代视觉去噪之间的结构不匹配。我们发现自回归归一化流是自回归Transformer，共享相同的因果掩码、KV缓存机制和从左到右的结构，使其成为真正统一多模态生成的最自然范式。我们提出了STARFlow2，基于Pretzel架构，通过残差跳跃连接垂直交织预训练VLM流与TarFlow流，两者均在相同的因果掩码下运行。结合深度浅层流设计和统一FAE潜在空间，STARFlow2实现了缓存友好的交织生成，其中文本和视觉输出可直接进入KV缓存而无需重新编码。实验表明，STARFlow2在图像生成和多模态理解基准上表现出色，验证了自回归流作为统一多模态建模基础的有效性。
英文摘要：
Deep generative models have advanced rapidly across text and vision, motivating unified multimodal systems that can understand, reason over, and generate interleaved text-image sequences. Most existing approaches combine autoregressive language modeling with diffusion-based image generators, inheriting a structural mismatch between causal text generation and iterative visual denoising. We observe that autoregressive normalizing flows are autoregressive Transformers--sharing the same causal mask, KV-cache mechanism, and left-to-right structure as LLMs--making them the most natural paradigm for true unified multimodal generation. We present STARFlow2, built on the Pretzel architecture that vertically interleaves a pretrained VLM stream with a TarFlow stream via residual skip connections, both operating under the same causal mask. Combined with a deep-shallow flow design and a unified FAE latent space, STARFlow2 enables cache-friendly interleaved generation where both text and visual outputs directly enter the KV-cache without re-encoding. Experiments demonstrate strong performance across image generation and multimodal understanding benchmarks, validating autoregressive flows as a viable foundation for unified multimodal modeling.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
