This repository provides source code for a critical analysis of deep learning progress in recommender systems research. It implements various neural recommendation algorithms including BPRMF, NeuMF, EASE, and other matrix factorization and embedding-based approaches. The framework enables reproducibility of published findings and comparison of neural versus traditional recommendation methods on standard benchmarks.