Convert varied-sparse vectors to fixed-sparse vectors?

is there an algorithm to convert varied-sparse vectors to fixed-sparse vectors … where the input vector can be one active bit to full 1s

The full range 2^n vectors can be used, but in practice only small portion get used. Which ones is not known in advance.

SPooler works only on fixed sparsity for input and output vectors , right ?

SP does not require fixed sparsity for its input.

1 Like

so it will work even if i have 1bit/1000, 2bits/2000, … vectors ? cool