{"version":"1.0","type":"rich","provider_name":"gaks.ai AI Glossary","provider_url":"https://gaks.ai/glossary","title":"Sparse Autoencoder — AI Glossary","author_name":"Glenn Katrud Solheim","author_url":"https://gaks.ai","width":600,"height":200,"html":"

AI Glossary — gaks.ai

Sparse Autoencoder

A type of autoencoder trained to produce representations where most activations are zero, with only a small number of features active for any given input. Sparse autoencoders have become a central tool in mechanistic interpretability research, used to decompose the dense, polysemantic activations of large language models into more interpretable features that correspond to specific, human-understandable concepts.

Source: gaks.ai/glossary/sparse-autoencoder →
"}