Over the past decade there has been an increasing amount of evidence that high resolution numerical methods for hyperbolic partial differential equations have an embedded (or "implicit") turbulence model. The present chapter describes this general class of methods and outlines the basic structure of high resolution methods as an effective turbulence model in the context of large eddy simulation (LES). This discussion is an extension of the MILES concept introduced by Boris, where monotone numerical algorithms are used for LES (MILES is an acronym for monotone integrated LES). We show that the implicit modeling includes elements of nonlinear eddy viscosity, scale-similarity and an effective dynamic model. In addition, we give examples of both success and failures with currently available methods and examine the effects of the embedded modeling in contrast to widely used subgrid scale (SGS) models.