YES

TRS:

  filter(cons(X),0(),M) -> cons(0())
  filter(cons(X),s(N),M) -> cons(X)
  sieve(cons(0())) -> cons(0())
  sieve(cons(s(N))) -> cons(s(N))
  nats(N) -> cons(N)
  zprimes() -> sieve(nats(s(s(0()))))

max/plus interpretations on N:

  filter_A(x1,x2,x3) = max{5, 4 + x1, 5 + x2, 5 + x3}
  filter#_A(x1,x2,x3) = max{5, 4 + x1, 5 + x2, 5 + x3}
  cons_A(x1) = max{4, 1 + x1}
  cons#_A(x1) = max{4, 1 + x1}
  0_A = 2
  0#_A = 2
  s_A(x1) = max{0, x1}
  s#_A(x1) = max{0, x1}
  sieve_A(x1) = max{1, x1}
  sieve#_A(x1) = max{1, x1}
  nats_A(x1) = max{5, 5 + x1}
  nats#_A(x1) = max{5, 5 + x1}
  zprimes_A = 8
  zprimes#_A = 8

precedence:

  filter > cons = zprimes > sieve = nats > 0 = s