Metraipolis › archive › dc_b87ff4df2cd0

Creative Generation via Multi-Agent Debate: Does Debate Suppress Diversity? — arXiv:2609.00683

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.00683
License (as archived): CC0-1.0
Archived: 2026-10-17T06:41:18.903Z
SHA-256 of the archived text: 74e74ac2e53ca7c0b7f512eb0302bc4185c84a3cb18ff69744c037e55493068e
Stable URL of this page: https://metraipolis.com/documento/dc_b87ff4df2cd0

How to cite

Creative Generation via Multi-Agent Debate: Does Debate Suppress Diversity? — arXiv:2609.00683 — arXiv (en), https://arxiv.org/abs/2609.00683 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b87ff4df2cd0 (Metraipolis archive dc_b87ff4df2cd0, sha256 74e74ac2e53ca7c0b7f512eb0302bc4185c84a3cb18ff69744c037e55493068e)

BibTeX

@misc{dcb87ff4df2cd0,
  title = {Creative Generation via Multi-Agent Debate: Does Debate Suppress Diversity? — arXiv:2609.00683},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.00683},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b87ff4df2cd0; sha256 74e74ac2e53ca7c0b7f512eb0302bc4185c84a3cb18ff69744c037e55493068e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00683. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Creative generation tasks, such as narrative writing and scientific ideation, demand both high-quality outputs and distinct responses across independent runs to maximize exploration. Multi-Agent Debate (MAD) has shown strong quality gains on factual and reasoning tasks, making it a natural candidate for creative generation. However, we find its convergence-driven design actively suppresses output diversity across independent runs, creating an inherent trade-off with creative tasks. We theoretically show that preserving diversity among agents within each debate session is a necessary condition for achieving diverse outputs across independent runs. Building on this finding, we propose Creative-MAD, which introduces two synergistic interventions to sustain agent divergence. Specifically, Cognitive Lens Assignment counters identity drift by anchoring each agent to a distinct and persistent cognitive mode, while Embedding-based Peer Selection counters majority pull by limiting each agent's context to its most semantically distant peers. Experiments across four creative benchmarks demonstrate that Creative-MAD significantly enhances both lexical and semantic diversity while maintaining MAD's output quality.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.00683 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T06:33:38.820Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).