crosscoders

Here are 3 public repositories matching this topic...

hematteo / learning-to-read-out

Code for "Learning to Read Out: Unembedding Dynamics in Language Model Pretraining" — parameter-trajectory crosscoders on the unembedding matrix W_U. Pretrained crosscoders: hf.co/hematteo/parameter-trajectory-crosscoders

language-models pythia sparse-autoencoders interpretability mechanistic-interpretability crosscoders pretraining-dynamics

Updated Jun 14, 2026
Python

teruyo14 / variant-specific-crosscoder-features

Star

Supplementary code & results for "Variant-specific crosscoder features are seed-stable but not detectably task-causal in a GRPO-LoRA math setting" (ICML 2026 Mech Interp Workshop, Spotlight)

reinforcement-learning sparse-autoencoders interpretability mechanistic-interpretability crosscoders model-diffing