A robotic chemistry policy reportedly made more than 3,000 decisions in 30 days
Discovery at Scale says its October 19 issue draws on 60 papers from the previous seven days.
TLDR
Discovery at Scale says a policy ran for 30 days and made more than 3,000 decisions on a robotic chemistry platform; a separate model predicted outcomes. The roundup also highlights research suggesting that text representations of crystals learn which elements are present far more readily than where the atoms sit.
Combined views
574
1 Source, first seen ago
