GradAttn: Replacing Fixed Residual Connections with Task-Modulated Attention Pathways
arXiv:2603.26756v1 Announce Type: new
Abstract: Deep ConvNets suffer from gradient signal degradation as network depth increases, limiting effective feature learning in complex architectures. ResNet addressed this through residual connections, but the…