潛移默化：將隱性推理融入圖像生成的藝術

摘要

文本到图像生成技术虽已取得显著进展，但现有方法普遍缺乏人类创造力特有的动态推理与优化能力。当前基于推理增强的范式大多依赖显性思维过程，需在固定步骤将中间推理解码为离散文本，并频繁进行图像编解码，导致效率低下、信息丢失及认知错配。为弥补这一缺陷，我们提出LatentMorph——一种将隐式潜在推理无缝集成到T2I生成过程的新框架。其核心包含四个轻量级组件：（一）用于将中间生成状态压缩为紧凑视觉记忆的冷凝器；（二）将潜在思维转化为可操作指引的转换器；（三）动态引导后续图像令牌预测的塑形器；（四）通过强化学习训练、自适应决定推理时机的调用器。通过在连续潜在空间中完成推理，LatentMorph规避了显式推理的瓶颈，实现了更高效的自适应优化。大量实验表明：LatentMorph（I）在GenEval和T2I-CompBench基准上分别将基础模型Janus-Pro性能提升16%和25%；（II）在WISE和IPV-Txt等抽象推理任务中超越显式推理范式（如TwiG）15%和11%；（III）同时将推理时间减少44%，令牌消耗降低51%；（IV）在推理调用机制上展现出与人类直觉71%的认知一致性。

English

Text-to-image (T2I) generation has achieved remarkable progress, yet existing methods often lack the ability to dynamically reason and refine during generation--a hallmark of human creativity. Current reasoning-augmented paradigms most rely on explicit thought processes, where intermediate reasoning is decoded into discrete text at fixed steps with frequent image decoding and re-encoding, leading to inefficiencies, information loss, and cognitive mismatches. To bridge this gap, we introduce LatentMorph, a novel framework that seamlessly integrates implicit latent reasoning into the T2I generation process. At its core, LatentMorph introduces four lightweight components: (i) a condenser for summarizing intermediate generation states into compact visual memory, (ii) a translator for converting latent thoughts into actionable guidance, (iii) a shaper for dynamically steering next image token predictions, and (iv) an RL-trained invoker for adaptively determining when to invoke reasoning. By performing reasoning entirely in continuous latent spaces, LatentMorph avoids the bottlenecks of explicit reasoning and enables more adaptive self-refinement. Extensive experiments demonstrate that LatentMorph (I) enhances the base model Janus-Pro by 16% on GenEval and 25% on T2I-CompBench; (II) outperforms explicit paradigms (e.g., TwiG) by 15% and 11% on abstract reasoning tasks like WISE and IPV-Txt, (III) while reducing inference time by 44% and token consumption by 51%; and (IV) exhibits 71% cognitive alignment with human intuition on reasoning invocation.

潛移默化：將隱性推理融入圖像生成的藝術

Show, Don't Tell: Morphing Latent Reasoning into Image Generation

摘要

Support