Sparse Autoencoders Can Interpret Randomly Initialized Transformers

Authors: Thomas Heap, Tim Lawson, Lucy Farnik, Laurence Aitchison

License: CC BY 4.0

Abstract: Sparse autoencoders (SAEs) are an increasingly popular technique for interpreting the internal representations of transformers. In this paper, we apply SAEs to 'interpret' random transformers, i.e., transformers where the parameters are sampled IID from a Gaussian rather than trained on text data. We find that random and trained transformers produce similarly interpretable SAE latents, and we confirm this finding quantitatively using an open-source auto-interpretability pipeline. Further, we find that SAE quality metrics are broadly similar for random and trained transformers. We find that these results hold across model sizes and layers. We discuss a number of number interesting questions that this work raises for the use of SAEs and auto-interpretability in the context of mechanistic interpretability.

Submitted to arXiv on 29 Jan. 2025

Explore the paper tree

Click on the tree nodes to be redirected to a given paper and access their summaries and virtual assistant

Also access our AI generated Summaries, or ask questions about this paper to our AI assistant.

Look for similar papers (in beta version)

By clicking on the button above, our algorithm will scan all papers in our database to find the closest based on the contents of the full papers and not just on metadata. Please note that it only works for papers that we have generated summaries for and you can rerun it from time to time to get a more accurate result while our database grows.