Pre-extracted OpenLORIS EfficientNet-B0 features for CLP-SNN

Pre-extracted EfficientNet-B0 (ImageNet-pretrained) feature vectors for the OpenLORIS 40-class continual learning for object-recognition benchmark, used in the paper "Online Continual Learning on Intel Loihi 2 via a Co-designed Spiking Neural Network." (Hajizada et al. 2026). Each feature is a 1280-dimensional L2-normalized vector. Contents (unzip at the repository root of github.com/elvinhajizada/CLP-SNN so the layout matches `data/...`): data/1shot/X_train_1_shot_{10,20,30}.pt — 1-shot training sets (3 seeds)data/1shot/y_train_1_shot_{10,20,30}.pt — labelsdata/25shot/X_train_25_shot_{10,20,30}.npy — 25-shot training sets (3 seeds)data/25shot/y_train_25_shot_{10,20,30}.npy — labelsdata/X_test.npy — balanced test set (60/class)data/y_test.npy — test labels The raw OpenLORIS-Scene videos are distributed by the OpenLORIS project at https://lifelong-robotic-vision.github.io/dataset/scene.html — this record provides only the pre-extracted feature embeddings used to reproduce the results in the paper.