In mammals, early embryonic development is implemented through the continuous flux of morphological and cellular state changes that specify the basic embryonic lineages (gastrulation) and later form the first organs (organogenesis). Yet, how variation is achieved while basic units (cells) in this system comprise identical genetic information, represents a deep and fundamental open question in biology. Recent developments in single-cell transcription and epigenome technologies are transforming our ability to capture the emergence of embryonic cellular diversification at the native state, and ultimate resolution. But, we believe that an urgent need in the field is to match descriptive single cell atlases with models and methodologies to derive novel understanding of function and regulation of embryonic tissues. In my talk, I will discuss our approach to defining embryonic gene function based on time-resolved models of mouse gastrulation and classical chimera assays that control for temporal, lineage, and cell-intrinsic/inter-cellular effects. As genes of interest, we focus on a particularly challenging and important example of the Dnmt and Tet family of DNA methylation and demethylation enzymes. From this starting point, we aspire to carefully work out epigenetic mechanisms by which cis-regulatory elements provide embryonic cells with robust identities and functions.