Hands-on Multi-agent Task Decomposition Experimentation

Teams will design and test decompositions, then study where information is lost, duplicated, misunderstood, or poorly integrated across conversations.