可解释且可引导的概念瓶颈稀疏自编码器
Interpretable and Steerable Concept Bottleneck Sparse Autoencoders
为视觉语言模型与图像生成模型提供了一种事后可插拔的稀疏自编码器改造方案：先剪除低效用神经元，再用对齐用户概念集的轻量概念瓶颈扩充潜空间，从而让潜变量既可解释又可被定向操控，可直接用于扩散模型的概念级编辑与生成控制。
原文依据 we propose Concept Bottleneck Sparse Autoencoders (CB-SAE)--a novel post-hoc framework that prunes low-utility neurons and augments the latent space
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论稀疏自编码器（SAE）有望为大型语言模型与大型视觉语言模型中的机制可解释性、概念发现与模型引导提供统一途径。然而，要真正发挥这一潜力，所学到的特征必须同时具备可解释性与可引导性。为此，我们提出两个计算开销很小的可解释性与可引导性新指标，用于对大型视觉语言模型的稀疏自编码器进行系统分析。该分析揭示了两点观察：(i) 大多数稀疏自编码器神经元在可解释性或可引导性上表现偏低，或两者兼低，因而在下游使用中效果不佳；(ii) 用户期望的概念常常在稀疏自编码器中缺失，从而限制了其实际效用。针对这些局限，我们提出概念瓶颈稀疏自编码器（CB-SAE）——一种新的事后框架，它剪除低效用神经元，并用一个与用户定义概念集对齐的轻量概念瓶颈来扩充潜空间。所得的 CB-SAE 在多个大型视觉语言模型与图像生成任务上，将可解释性提升 32.1%，可引导性提升 14.5%。
英文原文摘要 1042 字符
Sparse autoencoders (SAEs) promise a unified approach for mechanistic interpretability, concept discovery, and model steering in LLMs and LVLMs. However, realizing this potential requires learned features to be both interpretable and steerable. To that end, we introduce two new computationally inexpensive interpretability and steerability metrics for a systematic analysis of LVLM SAEs. This uncovers two observations; (i) a majority of SAE neurons exhibit either low interpretability or low steerability or both, rendering them ineffective for downstream use; and (ii) user-desired concepts are often absent in the SAE, thus limiting their practical utility. To address these limitations, we propose Concept Bottleneck Sparse Autoencoders (CB-SAE)--a novel post-hoc framework that prunes low-utility neurons and augments the latent space with a lightweight concept bottleneck aligned to a user-defined concept set. The resulting CB-SAE improves interpretability by +32.1% and steerability by +14.5% across LVLMs and image generation tasks.