1: \begin{abstract}\

2: A modification of the Adaptive Biasing Force  method is introduced, in which the free energy is approximated by a sum of tensor products of one-dimensional functions. This  enables to handle a larger number of reaction coordinates than the classical algorithm. We prove the algorithm is well-defined and prove the long-time convergence toward a regularized version of the free energy for an idealized version of the algorithm. Numerical experiments demonstrate that the method is able to capture correlations between reaction coordinates.

3: 

4: \medskip

5: 

6: \noindent\textbf{keywords:} Monte Carlo methods ; tensor ; free energy ; importance sampling ; molecular dynamics.

7: 

8: \noindent\textbf{MSC class (2010):} 65C05 ; 65N12.

9: \end{abstract}

10: