陈柏江
retrieval, ranking, reinforcement learning, and the systems details that make models behave in production.
A compact note on training retrievers against the failures their own index exposes.
A short refresher on random surfers, sparse matrices, and the operational shape of PageRank.