Abstract: IceCube DeepCore is an extension of the IceCube Neutrino Observatory designed to measure GeV scale atmospheric neutrino interactions. Neutrino reconstruction and classification tasks are especially difficult at GeV scale energies in IceCube DeepCore due to sparse instrumentation. Convolutional neural networks (CNNs) have been found to have better success at neutrino event reconstruction than likelihood-based methods. I present a new CNN model that exploits time and depth translational symmetry in IceCube DeepCore data and present the model’s performance, specifically for neutrino flavor identification. This new CNN model will be used for inelasticity reconstruction and/or matter effect dependent neutrino oscillation studies.