The underlying interactions in physical and biological systems often lead to a variety of behaviors and emergent states or phases. Under certain conditions, these phases can be transformed from one to another. The phase transition behaviors can be described by the bifurcation or catastrophe where different stable/unstable states can branch out or meet together with the birth of the new and death of the old states. Despite significant efforts, how the bifurcation and catastrophe actually occur dynamically and the associated mechanisms for nonequilibrium systems are still not very clear. As an example, we study the underlying mechanism of cell differentiation through bifurcations. Cell differentiation is one of the key fate decision-making processes that a cell faces. It is crucial for the development of multicellular organisms. Under induction, gene regulation changes, or stochastic fluctuations, the cell fate decision-making processes can exhibit different types of bifurcations or phase transitions. In order to understand the underlying mechanism, it is crucial to find out where and how the bifurcation occurs. However, this is still largely unknown. In this study, we found that the average of the curl flux as a major component of the driving force for the dynamics in addition to the landscape gradient and the associated entropy production rate both reach maximum near the bifurcation. This indicates that the curl flux and entropy production rate may provide the dynamical and thermodynamic origins of the bifurcation/catastrophe or phase transitions for cell differentiation and this possibly applies to many other nonequilibrium active systems.