How sharing activations can reduce inference cost of neural fields by up to 300x.
A brief introduction to my ICLR 24 paper
[about]
[projects]
[blogs]
[cv]
Coming soon...