1 paper
Arnav Kundu, Prateeth Nayak, Priyanka Padmanabhan +1
Always-on machine learning models require a very low memory and compute footprint. Their restricted parameter count limits the model's capacity to learn, and the effectiveness of t…