transformer-lens

Here are 2 public repositories matching this topic...

We study whether categorical refusal tokens enable controllable and interpretable safety behavior in language models.

ai pytorch safety llama llm mechanistic-interpretability llm-safety refusal llama3 transformer-lens llm-refusal

Code used for reverse-engineering a “Query-Gated Courier” circuit in Gemma-2-2B for role-gated retrieval.

Add a description, image, and links to the transformer-lens topic page so that developers can more easily learn about it.

To associate your repository with the transformer-lens topic, visit your repo's landing page and select "manage topics."