Gemma Scope: helping the safety community shed light on the inner workings of language models » Insights-Where Knowledge Drives Success

Gemma Scope: helping the safety community shed light on the inner workings of language models

-

July 17, 2025

Announcing a comprehensive, open suite of sparse autoencoders for language model interpretability.

Source: https://deepmind.google/discover/blog/gemma-scope-helping-the-safety-community-shed-light-on-the-inner-workings-of-language-models/