The growing channel count of silicon probes has substantially increased the number of neurons recorded in electrophysiology (ephys) experiments, rendering traditional manual spike sorting impractical. Instead, modern ephys recordings are processed with automated methods that use waveform template matching to isolate putative single neurons. While scalable, automated methods rely on assumptions that often fail to account for biophysical changes in action potential waveforms, leading to systematic oversplitting of individual neurons into multiple putative units. Consequently, manual curation of these errors, which is both time-consuming and lacking in reproducibility, remains necessary. To improve efficiency and reproducibility in the spike-sorting pipeline, we introduce the Spike-sorting Lapse Amelioration System (SLAy), an algorithm that automatically merges oversplit spike units. SLAy employs two novel metrics: (1) a waveform similarity metric that uses a neural network to obtain spatially informed, nonlinear waveform representations, and (2) a cross-correlogram significance metric based on the earth mover’s distance between the observed and null cross-correlograms. To improve reproducibility and remove the need for manual tuning, we also develop an automatic parameter setting procedure for SLAy that accounts for dataset-specific characteristics. On simulated oversplitting across a diverse set of animal models, brain regions, and probe geometries, SLAy substantially outperforms an existing merging algorithm, achieving high recall without merging extraneous units. On the original datasets without simulated oversplitting, SLAy recovers ∼ 95% of merges found by human curators and human curators agree with ∼ 90% of merges suggested by SLAy. SLAy leverages multithreading for computational efficiency, running in less than 10 minutes for all recordings we tested. SLAy is also compatible with SpikeInterface, making it a practical and flexible solution for large-scale ephys data analysis across acquisition systems and spike sorters.