Learning Input Encodings for Kernel-Optimal Implicit Neural Representations | Read Paper on Bytez