1: \begin{abstract}

2: 	Iterative regularization exploits the implicit bias of an optimization algorithm to regularize ill-posed problems.

3:     Constructing algorithms with such built-in regularization mechanisms is a classic challenge in inverse problems but also  in modern machine learning, where it provides both a new perspective on algorithms analysis, and significant speed-ups compared to explicit regularization.

4: 	In this work, we propose and study the first iterative regularization procedure able to handle biases described by non smooth and non strongly convex functionals,  prominent in low-complexity regularization.

5: 	Our approach is based on a primal-dual algorithm  of which we analyze convergence and stability properties, even in the case where the original problem is unfeasible. The general results are illustrated considering the special case

6:  of sparse recovery with the $\ell_1$ penalty.

7: 	Our theoretical results are complemented by experiments showing the computational benefits of our approach.

8: \end{abstract}

9: