conftrace_

Or David Shafran

1 papers · 2026–2026 · 1 conference · across top CS/AI conferences

Conferences

ACL (1)

Top co-authors

Mor Geva (1) Atticus Geiger (1)

Keywords

mechanistic interpretability (1) sparse autoencoder (1) interpretable feature (1) large language model (1) semi-nonnegative matrix factorization (1) neuron group (1)

Papers

Constructing Interpretable Features from Compositional Neuron Groups ACL 2026