1: \begin{abstract}

2:     Being able to study the dynamics of quantum systems interacting with several environments is important in many settings ranging from quantum chemistry to quantum thermodynamics, through out-of-equilibrium systems. 

3:     For such problems tensor network-based methods are state-of-the-art approaches to perform numerically exact simulations. 

4:     However, to be used efficiently in this multi-environment non-perturbative context, these methods require a clever choice of the topology of the wave-function Ans\"atze. This is often done by analysing cross-correlations between different system and environment degrees of freedom.

5:     We show for canonical model Hamiltonians that simple orderings of bosonic environmental modes, which enable to write the joint \{System + Environments\} state as a matrix product state, reduce considerably the bond dimension required for convergence.

6:     These results suggest that complex correlation analyses in order to tweak tensor networks topology (e.g. entanglement renormalization) are usually not necessary and that tree tensor network states are sub-optimal compared to simple matrix product states in several applications.

7: \end{abstract}

8: