Large-deformation processes in solids are known to undermine the predictive capability of the standard finite element formulation when excessive mesh distortion occurs. Remeshing is commonly adopted to restore mesh quality; however, in inelastic analyses this strategy involves the remapping of internal variables from the discarded mesh to the new one, therefore inducing diffusion of the internal variables describing the material history, affecting accuracy. The main objective of this work is to assess the effectiveness of a mixed finite element formulation in reducing the diffusion of internal variables induced by remeshing. Building on earlier developments of [1], we formulate a variational framework for a saddle-point problem specifically tailored to large-deformation kinematics, together with a finite element discretization based on tetrahedral meshes. The resulting discrete variational statement naturally leads to a numerical approach in which the quadrature point locations coincide with those of the nodes, yielding a formulation that behaves as a nodally integrated finite element method, equivalent to the one proposed in [2], originally introduced to alleviate locking phenomena. Linking the material history with the nodes aligns naturally with remeshing strategies based on reconnecting existing nodes, thereby facilitating the transfer of state variables across successive meshes. In this context, conceptual similarities with particle-based approaches such as the smoothed particle finite element method [3] are identified, and the distinguishing features and advantages of the present formulation are discussed. Quantitative indicators to evaluate internal variable diffusion are introduced and employed in numerical simulations of large-strain elasto-plasticity. The results demonstrate the proposed formulation’s ability to reduce remapping-induced diffusion when compared to standard finite element implementations, highlighting its potential for problems involving severe deformation and frequent remeshing.