ReactNexus: Generative discovery of chemical reaction networks
ReactNexus explores chemical reaction networks from reactant graphs, without requiring predefined products or reaction templates.
Thoughts, stories and ideas.
ReactNexus explores chemical reaction networks from reactant graphs, without requiring predefined products or reaction templates.
This perspective proposes evaluating AI scientists through complete discovery episodes, including evidence, failures and null results.
MIRA coordinates specialized agents to connect hypotheses, simulations and automated experiments while preserving evidence and research memory.
Materials Property Axiom (MPA) uses multi-phase training to scale foundation models for materials science research.
SDE evaluates scientific reasoning through questions and research projects across biology, chemistry, materials science and physics.
SAGA pairs evolving scientific objectives with candidate optimization to explore research goals and their trade-offs.