Reproduce “Amora: Communication-Efficient Private Approximate Nearest Neighbors Search” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Qiuhao Wang, Zuowei Chen, Xu Yang, Saiyu Qi, Yong Qi
Amora reduces communication in private approximate nearest-neighbor search by separating vector distances from graph neighbor lists. It uses partition-level private retrieval to obtain distances for many vertices, then retrieves lists for selected groups. On AGNEWS, it achieves Recall@10 of 0.936 ± 0.001 with 0.89 MB per query, versus Pacmann’s 0.911 ± 0.011 and 82.88 MB; latency is not always lower.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.