Reproduce “Atom-Wise Allocation of Message-Passing Scale for Molecular Property Prediction” 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

Regina Qianru Zhang, Honggang Wen, Zongru Li, Jinsong Ouyang, Zhiqin Chu, Pietro Liò, Xiaofeng Liu, Siu-Ming Yiu

PRISM lets each atom in a molecular graph neural network (GNN) combine retained post-update states using a bounded scalar from its encoded features. A smooth unimodal profile shifts across depths, optionally alongside radius-matched fingerprint environments. On FreeSolv, attaching PRISM to a GCN lowers RMSE from 0.951 to 0.673; results vary by task, and computation still evaluates every bank state.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.