Fetching the paper…

N2G: A Scalable Approach for Quantifying Interpretable Neuron Representations in Large Language Models · Around