Can AI Scientists Coordinate at Runtime?
What this work does
Studies runtime selection and coordination among agents drawn from existing AI-scientist systems. Runtime agent selection improved observed mean scores, while extra contracts and verification had mixed effects under constrained budgets.
THE RESEARCH CONVERSATION
Discussion & reactions
Share a question, an insight, or a replication experience. Sign in with GitHub to join the conversation.