1: \begin{abstract}

2:   In this paper, we propose new accelerated methods for smooth convex

3:   optimization, called contracting proximal methods.

4:   At every step of these methods, we need to minimize

5:   a contracted version of the objective function augmented

6:   by a regularization term in the form of

7:   Bregman divergence.

8:   We provide global convergence analysis for a general scheme

9:   admitting inexactness in solving the auxiliary subproblem.

10:   In the case of using for this purpose high-order tensor methods,

11:   we demonstrate an acceleration effect

12:   for both convex and uniformly convex composite objective

13:   functions. Thus, our construction explains acceleration for methods of any order

14:   starting from one. The augmentation of the number of calls of

15:   oracle due to computing the contracted proximal steps is

16:   limited by the logarithmic factor in the worst-case complexity

17:   bound.

18: \end{abstract}

19: