This preprint was accepted December 22, 2006
ABSTRACT: We prove a time hierarchy theorem for inverting functions computable in a slightly non-uniform polynomial time. In particular, we prove that if there is a strongly one-way function then for any $k$ and for any polynomial $p$, there is a function $f$ computable in linear time with one bit of advice such that there is a polynomial-time probabilistic adversary that inverts $f$ with probability $\ge 1/p(n)$ on infinitely many lengths of input while all probabilistic $O(n^k)$-time adversaries with logarithmic advice invert $f$ with probability less than $1/p(n)$ on almost all lengths of input. We also prove a similar theorem in the worst-case setting, i.e., if $\P\neq\NP$, then for every $l>k\ge1$ $$ (\DTime{n^k} \cap \NTime{n})/1 \subsetneq (\DTime{n^l} \cap \NTime{n})/1. $$[Full text: (.ps.gz)]