DreamReader: 一种用于文本到图像模型的可解释性工具包
DreamReader: An Interpretability Toolkit for Text-to-Image Models
AI总结：
DreamReader提出了一种统一框架，通过可组合的表示操作分析扩散模型的可解释性，引入三种新干预机制，通过实验展示其在文本到图像模型中的应用，推动可解释性研究。
摘要：
尽管文本到图像（T2I）扩散模型被广泛应用，但因果和表示层面的分析仍碎片化且局限于孤立的探测技术。为解决这一差距，我们引入DreamReader：一种统一框架，将扩散可解释性形式化为可组合的表示操作，涵盖激活提取、因果补丁、结构化消融和激活引导，跨模块和时间步。DreamReader提供了一个模型无关的抽象层，使扩散架构的系统分析和干预成为可能。除了整合现有方法，DreamReader引入了三种新的干预原语：（1）表示微调（LoReFT）用于子空间约束的内部适应；（2）使用MLP探测器训练的分类引导梯度引导；（3）组件级跨模型映射，用于系统研究表示在不同模态间的可转移性。这些机制使我们能够通过借鉴语言模型可解释性技术，对T2I模型进行轻量级的白盒干预。我们通过受控实验展示DreamReader：（i）在两个模型之间执行激活拼接，（ii）应用LoReFT引导多个激活单元，可靠地将目标概念注入生成图像中。实验通过声明性方式指定，并在受控的批量管道中执行，以实现可重复的大规模分析。在多个案例研究中，我们展示了从语言模型可解释性技术适应到扩散模型的有希望且可控的干预。DreamReader作为开源工具包发布，以推动T2I可解释性研究。
英文摘要：
Despite the rapid adoption of text-to-image (T2I) diffusion models, causal and representation-level analysis remains fragmented and largely limited to isolated probing techniques. To address this gap, we introduce DreamReader: a unified framework that formalizes diffusion interpretability as composable representation operators spanning activation extraction, causal patching, structured ablations, and activation steering across modules and timesteps. DreamReader provides a model-agnostic abstraction layer enabling systematic analysis and intervention across diffusion architectures. Beyond consolidating existing methods, DreamReader introduces three novel intervention primitives for diffusion models: (1) representation fine-tuning (LoReFT) for subspace-constrained internal adaptation; (2) classifier-guided gradient steering using MLP probes trained on activations; and (3) component-level cross-model mapping for systematic study of transferability of representations across modalities. These mechanisms allows us to do lightweight white-box interventions on T2I models by drawing inspiration from interpretability techniques on LLMs. We demonstrate DreamReader through controlled experiments that (i) perform activation stitching between two models, and (ii) apply LoReFT to steer multiple activation units, reliably injecting a target concept into the generated images. Experiments are specified declaratively and executed in controlled batched pipelines to enable reproducible large-scale analysis. Across multiple case studies, we show that techniques adapted from language model interpretability yield promising and controllable interventions in diffusion models. DreamReader is released as an open source toolkit for advancing research on T2I interpretability.
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。
