1: \begin{abstract}

2: Consider a generalized Elephant Random Walk in which the step is chosen

3: by selecting $k$ previous steps, with $k$ odd, then going in the

4: majority direction with probability $p$ and in the opposite direction

5: otherwise. In the $k=1$ case, the model is the original one, and

6: can be solved exactly by analogy with Friedman's urn. However, the

7: analogy cannot be extended to the $k>2$ case already. In this paper

8: we show how to treat the model for each $k$ by analogy with the more

9: general urn model of Hill Lane and Sudderth. Interestingly, for $k>2$

10: we find a critical dependence from the initial conditions above certain

11: values of the memory parameter $p$, and regions of convergence with

12: entropy that is sub-linear in the number of steps.

13: \end{abstract}

14: