Decomposing The Dark Matter of Sparse Autoencoders

Published: 30 Mar 2025, Last Modified: 30 Mar 2025Accepted by TMLREveryoneRevisionsBibTeXCC BY 4.0
Abstract: Sparse autoencoders (SAEs) are a promising technique for decomposing language model activations into interpretable linear features. However, current SAEs fall short of completely explaining model performance, resulting in ``dark matter'': unexplained variance in activations. This work investigates dark matter as an object of study in its own right. Surprisingly, we find that much of SAE dark matter---about half of the error vector itself and $>90\% $ of its norm---can be linearly predicted from the initial activation vector. Additionally, we find that the scaling behavior of SAE error norms at a per token level is remarkably predictable: larger SAEs mostly struggle to reconstruct the same contexts as smaller SAEs. We build on the linear representation hypothesis to propose models of activations that might lead to these observations, including postulating a new type of ``introduced error''; these insights imply that the part of the SAE error vector that cannot be linearly predicted (``nonlinear'' error) might be fundamentally different from the linearly predictable component. To validate this hypothesis, we empirically analyze nonlinear SAE error and show that 1) it contains fewer not yet learned features, 2) SAEs trained on it are quantitatively worse, 3) it helps predict SAE per-token scaling behavior, and 4) it is responsible for a proportional amount of the downstream increase in cross entropy loss when SAE activations are inserted into the model. Finally, we examine two methods to reduce nonlinear SAE error: inference time gradient pursuit, which leads to a very slight decrease in nonlinear error, and linear transformations from earlier layer SAE outputs, which leads to a larger reduction.
Submission Length: Regular submission (no more than 12 pages of main content)
Previous TMLR Submission Url: https://openreview.net/forum?id=vkwflBsZRL
Changes Since Last Submission: Fixing format by adding back line under header.
Code: https://github.com/JoshEngels/SAE-Dark-Matter
Assigned Action Editor: ~Chinmay_Hegde1
Submission Number: 3761
Loading

OpenReview is a long-term project to advance science through improved peer review with legal nonprofit status. We gratefully acknowledge the support of the OpenReview Sponsors. © 2025 OpenReview