paper
referenced-only
2022
paper:2022-transformerTransformer Circuits Thread
Similar preprints — Semantic Scholar
Cited by (2)
- Toy Models of Superposition
Neural networks store more sparse features than they have dimensions, and a one-line change shows it: adding a ReLU to the output of a linear W^T W autoencoder moves the optimum from PCA-like top-m fe
- Finding Alignments Between Interpretable Causal Variables and Distributed Neural Representations
Distributed alignment search (DAS) resolves two blocking limitations of prior causal abstraction work—brute-force alignment search and the localist assumption that high-level variables map to disjoint