Finite-Size Scaling in Random $K$-SAT Problems

ORAL

Abstract

We propose a comprehensive view of threshold behaviors in random $K$-satisfiability ($K$-SAT) problems, in the context of the finite-size scaling (FSS) concept of nonequilibrium absorbing phase transitions using the average SAT (ASAT) algorithm. In particular, we focus on the value of the FSS exponent to characterize the SAT/UNSAT phase transition, which is still debatable. We also discuss the role of the noise (temperature-like) parameter in stochastic local heuristic search algorithms.

Authors

  • Meesoon Ha

    Dept. of Physics, KAIST

  • Sang Hoon Lee

    Dept. of Physics, KAIST

  • Chanil Jeon

    Dept. of Physics, KAIST

  • Hawoong Jeong

    Dept. of Physics, KAIST, Department of Physics, Korea Advanced Institute of Science and Technology