paper-with-me

Papers

Toward Identifiable Sparse Autoencoders

2026-05-29 · Walter Nelson, Theofanis Karaletsos, Francesco Locatello arxiv

Recently, sparse autoencoders (SAEs) have emerged as an attractive tool for interpreting and interacting with representations in practical neural networks. While it is common empirical folklore, we also show theoretically that SAEs are highly unstable: different training runs are likely to produce different concept dictionaries and sparse codes. We characterize the model properties that hinder the stability of real-world SAEs, and address each of these problems through minimal changes to the architecture and training procedure. Together, these changes yield two versions of an \textbf{i}dentifiable SAE (iSAE), a variant of the standard TopK SAE with lower reconstruction error and improved stability. We explain this improvement theoretically by connecting SAEs with traditional dictionary learning approaches, and show that the dictionaries learned in practice satisfy an approximate restricted isometry condition, rendering the corresponding sparse codes in those models near-identifiable.

📄 PDF Abstract BibTeX arXiv:2605.31245

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Can sparse autoencoders make sense of latent representations?

2024-10-15 · Viktoria Schuster

Sparse autoencoders (SAEs) have lately been used to uncover interpretable latent features in large language models. Here, we explore their potential for decomposing latent representations in complex and high-dimensional …

Steering Vision-Language Models with Joint Sparse Autoencoders

2026-06-24 · Huizhen Shu, Xuying Li, Hongxu Lin, Wenjie Sun 외 arxiv

Sparse Autoencoders (SAEs) have shown promise for analyzing language models, but applying them to vision-language models (VLMs) often yields representations that are difficult to use as controllable cross-modal steering …

Interpreting Video Representations with Spatio-Temporal Sparse Autoencoders

2026-04-05 · Atahan Dokme, Sriram Vishwanath arxiv

We present the first systematic study of Sparse Autoencoders (SAEs) on video representations. Standard SAEs decompose video into interpretable, monosemantic features but destroy temporal coherence: hard TopK selection pr…

Action ClassificationVideo Retrieval

Interpretable GOHR Agents via Sparse Autoencoders

2026-07-27 · Shiwei Tan, Yusong Zhao, Weiyi Qin, Wentian Wang 외 arxiv

A central challenge in interpreting learned decision-making systems is to determine whether their internal representations contain concepts that help explain their behavior. We report interpretability experiments for a t…

Posterior Collapse and Latent Variable Non-identifiability

2023-01-02 · NeurIPS 2021 12 · Yixin Wang, David M. Blei, John P. Cunningham

Variational autoencoders model high-dimensional data by positing low-dimensional latent variables that are mapped through a flexible distribution parametrized by a neural network. Unfortunately, variational autoencoders …

AttributeVariational Inference