Reproduce “Searching the Pareto Front: Multi-Objective Retrieval for Heterogeneous Scientific Modalities” 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
Song Young Oh, Nesar Ramachandra, Bin Xia, Ian Foster, Kyle Chard
ParetoFuse retrieves simulated galaxies by keeping each modality’s distance separate and selecting candidates that are not dominated across modalities. It then ranks those candidates using distances in concatenated embeddings. On a Last Journey simulation, its single-neighbor result has lower average raw-feature spread across six properties than MOSAIC-lite, but MOSAIC-lite has lower aggregate spread at five neighbors; surveys were not tested.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.