This repository contains experiment examples demonstrating how to use the Steering API to study and manipulate model behavior through SAE (Sparse Autoencoder) features. - View it on GitHub