rmovva/HypotheSAEs

HypotheSAEs: hypothesizing interpretable relationships in text datasets using sparse autoencoders. https://arxiv.org/abs/2502.04382

语言：Jupyter Notebook · GitHub 星标：100 · Fork：32

话题：ai-for-science · computational-social-science · interpretability · nlp · sparse-autoencoders · topic-modeling

在 GitHub 上查看原仓库