Reproduce “NEST: Nested Embedding Representation for Multiscale Clustering” 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
Zhanqi Zhang, Mikio Aoi, Gal Mishne
NEST learns multiscale clustering in one representation by adjusting a frozen feature extractor and training longer embedding prefixes for finer groupings. It uses known cluster counts at each level, without parent-child labels or a ground-truth tree. On CIFAR-100, it reaches class-level ACC of .35±.01 versus .27±.01 for DeepECT, in the reported dataset and setup.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.