1: \begin{abstract}

2: 

3: We compare the convergence of several flat-histogram methods applied to the 2D

4: Ising model, including the recently introduced stochastic approximation with a

5: dynamic update factor (SAD) method. We compare this method with the Wang-Landau

6: (WL) method, the $1/t$ variant of the WL method, and standard stochastic

7: approximation Monte Carlo (SAMC). In addition, we consider a procedure WL

8: followed by a ``production run'' with fixed weights that refines the estimation

9: of the entropy. To our knowledge, this work is the first to test this approach

10: against other methods. We find that WL followed by a production run \emph{does}

11: converge to the true density of states, in contrast to pure WL. Three of the

12: methods converge robustly: SAD, $1/t$-WL, and WL followed by a production run.

13: Of these, SAD does not require \emph{a priori} knowledge of the energy range.

14: This work also shows that WL followed by a production run performs superior to

15: other forms of WL while ensuring both ergodicity and detailed balance.

16: 

17: \end{abstract}

18: