1: \begin{abstract}

2: We study the competition between interference

3: due to multiple single-particle paths and Coulomb interaction in a

4: simple model of an Anderson-like impurity with

5: local-magnetic-field-induced level splitting coupled to

6: ferromagnetic leads. The model along with its potential

7: experimental relevance in the field of spintronics serves as a

8: nontrivial benchmark system where various quantum transport

9: approaches can be tested and compared. We present results for the

10: linear conductance obtained by a spin-dependent implementation of

11: the density matrix renormalization group scheme which are compared

12: with a mean-field solution as well as a seemingly more advanced

13: Hubbard-I approximation. We explain why mean-field yields nearly

14: perfect results, while the more sophisticated Hubbard-I approach

15: fails, even at a purely conceptual level since it breaks

16: hermiticity of the related density matrix. Furthermore, we study

17: finite bias transport through the impurity by the mean-field

18: approach and recently developed higher-order density matrix

19: equations. We find that the mean-field solution fails to describe

20: the plausible results of the higher-order density matrix approach

21: both quantitatively and qualitatively as it does not capture some

22: essential features of the current-voltage characteristics such as

23: negative differential conductance.

24: \end{abstract}