Multi-agent debate architecture — proposer, critic, judge, rounds

Why architecture matters here

Debate can improve accuracy on hard reasoning but is expensive. Architecture matters because proposers + judge + rounds compose.

The architecture: every piece explained

The top strip is roles. Question. Proposer A. Proposer B. Critic.

The middle row is process. Rebuttal round. Judge. Final answer. Rounds cap.

The lower rows are ops. Bias risk. Metrics. Ops — persona + judge + termination.

End-to-end flow

End-to-end: question posed. A and B answer. Critic attacks both. Round 2: each defends or revises. Judge sees all and picks. Return winner with rationale.