1: \begin{abstract}

2: Traditional weak-lensing mass reconstruction techniques suffer from various artifacts, including noise amplification and the mass-sheet degeneracy. In Hong et al. (2021), we demonstrated that many of these pitfalls of traditional mass reconstruction can be mitigated using a deep learning approach based on a convolutional neural network (CNN). In this paper, we present our improvements and report on the detailed performance of our CNN algorithm applied to next-generation wide-field observations. Assuming the field of view ($3\fdg 5 \times 3\fdg5$) and depth (27 mag at $5\sigma$) of the Vera C. Rubin Observatory, we generated training datasets of mock shear catalogs with a source density of 33 arcmin$^{-2}$ from cosmological simulation ray-tracing data. We find that the current CNN method provides high-fidelity reconstructions consistent with the true convergence field, restoring both small and large-scale structures. In addition, the cluster detection utilizing our CNN reconstruction achieves $\sim75$\% completeness down to $\sim 10^{14}M_{\sun}$. We anticipate that this CNN-based mass reconstruction will be a powerful tool in the Rubin era, enabling fast and robust wide-field mass reconstructions on a routine basis.

3: \end{abstract}

4: