Perform mechanistic interpretability research
Evaluate the precise behavior of Gemma 2 models with layer-level analysis.
Gemma Scope
A set of interpretability tools built to help researchers understand the inner workings of Gemma 2.
Explore Gemma Scope
Examine the behavior of individual model layers, while the model processes requests — to help address critical concerns including hallucinations, biases, and manipulation.
Slide 1 of 2
Capabilities
Gemma Scope provides researchers with a suite of sparse autoencoders. Think of these as microscopes that let you zoom in on dense, compressed activations, and expand them to larger, sparser, more interpretable forms.
Debug model behavior
Pinpoint the source of specific model issues (such as biases and hallucinations) by examining layer-specific representations.
Download Gemma Scope
