Reproduce “Building heterogeneous multistep workflows in automated microscopy: reward design and knowledge transfer for autonomous structure property discovery” 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
Sergei V. Kalinin
The framework schedules imaging, spectroscopy, and, in a second campaign, recalibration under an acquisition budget. It uses simulated campaigns built on pre-acquired image–EELS measurements, with separate action costs and objectives for finding strong responses or learning structure–response relationships. For random forest, exploration had lower held-out RMSE (33.96 ± 1.32), while exploitation found higher mean best band area (250.27 ± 2.11); results are simulated, not live microscope control.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.