The impact of multi-agent debate protocols on debate quality: a controlled case study

Ramtin Zargari Marandi · 2026

Read The impact of multi-agent debate protocols on debate quality: a controlled case study by Ramtin Zargari Marandi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-28813.txt.

Description

Marandi studies how debate protocol choices affect quality in multi-agent debate systems. Marandi compares three protocols against a no-interaction baseline. Agents in one setup see only current-round contributions, while another gives full prior-round context. A new rank-based version reorders agents and silences one per round via a judge model. In a macroeconomic case study with 20 events, the rank-based one converged fastest. The no-interaction baseline gave the highest argument diversity. Protocol design clearly matters.

Catalog record

CREATORRamtin Zargari Marandi
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33036
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.28813

Browse the full library · About the Tower of Babel