AI RESEARCH

Neuron Populations Exhibit Divergent Selectivity with Scale

arXiv CS.LG

ArXi:2606.03990v1 Announce Type: new We investigate whether neuron populations within neural networks evolve predictably with scale, extending scaling laws beyond macroscopic observables such as loss. To probe this question, we study Rosetta Neurons, a previously characterized class of neurons whose activation patterns are similar across independently trained models (Dravid, 2023