1: \begin{abstract}

2: The Frank-Wolfe algorithm has regained much interest

3: in its use in structurally constrained machine learning applications. However, one major limitation of the Frank-Wolfe algorithm is the slow local convergence property due to the zig-zagging behavior.

4: We observe the zig-zagging phenomenon in the Frank-Wolfe method as an artifact of discretization,

5: and propose multistep Frank-Wolfe variants 

6: where the truncation errors decay as $O(\Delta^p)$, where $p$ is the method's order. This strategy ``stabilizes" the method, and allows tools like line search and momentum to have more benefit. However, our results suggest that the worst case convergence rate of Runge-Kutta-type discretization schemes cannot improve upon that of the vanilla Frank-Wolfe method for a rate depending on $k$. Still, we believe that this analysis adds to the growing knowledge of flow analysis for optimization methods, and is a cautionary tale on the ultimate usefulness of multistep methods.

7: \end{abstract}

8: