Single-cell electrophysiological recordings provide a powerful window into neuronal functional diversity and offer an interpretable route for linking intrinsic physiology to transcriptomic identity. Here, we adapt the IPFX/Gouwens electrophysiological feature-family organization introduced by Gouwens et al. (2020) using publicly available Allen Institute Patch-seq datasets from mouse and human cortex. We focused on GABAergic inhibitory interneurons in an aligned subclass structure (Lamp5, Pvalb, Sst, Vip). After quality control, we analyzed 3,699 mouse visual cortex neurons and 506 human neocortical neurons from neurosurgical resections. Using standardized electrophysiological features and a Gouwens-style sparse-PCA random-forest baseline, we recovered major class-level separations consistent with the original mouse study. For supervised prediction, a class-balanced random forest provided a strong feature-engineered baseline in mouse data and a reduced but still informative baseline in human data. We then developed an attention-based BiLSTM that deliberately diverges from the sparse-PCA baseline by operating directly on the pre-sPCA family-vector tensor rather than on the 44 sparse-PC representation, providing feature-family-level interpretability via learned attention weights. Finally, we evaluated cross-species transfer by pretraining on mouse data and fine-tuning on human data for an aligned 4-class task, improving human macro-F1 versus human-only training. Together, these results support the utility of the Gouwens feature-family organization while distinguishing the sPCA-compressed baseline from our family-structured neural sequence model.
