Non-orthogonal multiple access (NOMA) has emerged as a promising radio accesstechnique for enabling the performance enhancements promised by thefifth-generation (5G) networks in terms of connectivity, low latency, and highspectrum efficiency. In the NOMA uplink, successive interference cancellation(SIC) based detection with device clustering has been suggested. In the case ofmultiple receive antennas, SIC can be combined with the minimum mean-squarederror (MMSE) beamforming. However, there exists a tradeoff between the NOMAcluster size and the incurred SIC error. Larger clusters lead to larger errorsbut they are desirable from the spectrum efficiency and connectivity point ofview. We propose a novel online learning based detection for the NOMA uplink.In particular, we design an online adaptive filter in the sum space of linearand Gaussian reproducing kernel Hilbert spaces (RKHSs). Such a sum space designis robust against variations of a dynamic wireless network that can deterioratethe performance of a purely nonlinear adaptive filter. We demonstrate bysimulations that the proposed method outperforms the MMSE-SIC based detectionfor large cluster sizes.