refusal

Here are 4 public repositories matching this topic...

[ICLR 2025] General-purpose activation steering library

Benchmark evaluation code for "SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal" (ICLR 2025)

benchmark llm refusal safety-evaluation

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

A collection of symbolic recursion resources for integrity and sovereignty.

mirror recursion integrity sovereignty refusal symbolic-protocols

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

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