SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability

12 March 2025

Papers citing "SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability"

7 / 7 papers shown

Title
Evaluating Explanations: An Explanatory Virtues Framework for Mechanistic Interpretability -- The Strange Science Part I.ii Kola Ayonrinde Louis Jaburi XAI 71 1 0 02 May 2025
MIB: A Mechanistic Interpretability Benchmark Aaron Mueller Atticus Geiger Sarah Wiegreffe Dana Arad Iván Arcuschin ... Alessandro Stolfo Martin Tutek Amir Zur David Bau Yonatan Belinkov 41 1 0 17 Apr 2025
$SAEs $\textit{Can}$ Improve Unlearning: Dynamic Sparse Autoencoder Guardrails for Precision Unlearning in LLMs$ SAEs $\textit{Can}$ Improve Unlearning: Dynamic Sparse Autoencoder Guardrails for Precision Unlearning in LLMs Aashiq Muhamed Jacopo Bonato Mona Diab Virginia Smith MU 58 0 0 11 Apr 2025
Evaluating and Designing Sparse Autoencoders by Approximating Quasi-Orthogonality Sewoong Lee Adam Davies Marc E. Canby J. Hockenmaier LLMSV 65 0 0 31 Mar 2025
Revisiting End-To-End Sparse Autoencoder Training: A Short Finetune Is All You Need Adam Karvonen 34 0 0 21 Mar 2025
Learning Multi-Level Features with Matryoshka Sparse Autoencoders Bart Bussmann Noa Nabeshima Adam Karvonen Neel Nanda 54 0 0 21 Mar 2025
Rethinking Evaluation of Sparse Autoencoders through the Representation of Polysemous Words Gouki Minegishi Hiroki Furuta Yusuke Iwasawa Y. Matsuo 49 1 0 09 Jan 2025