10:35
2026-05-26
letsdatascience.com
large-language-models
Sparse Autoencoders Reveal Cortical Brain-LLM Semantic Mapping
Researchers at arXiv (arXiv:2605.23035) used sparse autoencoders to decompose GPT-2 XL and Llama-3.1-8B into 16K-32K interpretable features per layer, finding that semantic features alone recovered 94…