Metraipolis › archive › dc_5fb0c0787f6a
Multi-agent Scaling Across Disjunctive and Compensatory Tasks — arXiv:2609.31563
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.31563
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T01:37:47.082Z
- SHA-256 of the archived text
e42d00d75fbe58c02e7905d1667ff6c31d636897652ab0774e3d5132b083d996- Stable URL of this page
https://metraipolis.com/documento/dc_5fb0c0787f6a
How to cite
Multi-agent Scaling Across Disjunctive and Compensatory Tasks — arXiv:2609.31563 — arXiv (en), https://arxiv.org/abs/2609.31563 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5fb0c0787f6a (Metraipolis archive dc_5fb0c0787f6a, sha256 e42d00d75fbe58c02e7905d1667ff6c31d636897652ab0774e3d5132b083d996)
BibTeX
@misc{dc5fb0c0787f6a,
title = {Multi-agent Scaling Across Disjunctive and Compensatory Tasks — arXiv:2609.31563},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.31563},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5fb0c0787f6a; sha256 e42d00d75fbe58c02e7905d1667ff6c31d636897652ab0774e3d5132b083d996},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.31563. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Multi-agent LLM systems are often expected to improve as team size increases, yet the scaling behavior may depend on task structure. Our central contribution is to introduce Steiner's taxonomy of group tasks as a framework for analyzing multi-agent LLM scaling and focusing the analysis on disjunctive and compensatory tasks. We model independently sampled agents as conditionally independent given the item, which yields their large-team limits: plurality voting converges to the model's modal answer, and averaging converges to the model's item-level bias. Across selected representative benchmarks, 13 open-weight models, and teams of up to 30 agents, we find qualitatively different scaling behavior. On disjunctive tasks, the probability that at least one agent is correct grows by 5-20 points with team size, but plurality voting over agents that answer directly realises almost none of this potential, as the model predicts to within 0.5 points on average. Multi-round revision raises accuracy considerably, yet the gain is nearly the same with one peer as with 29. In contrast, scaling provides little benefit on Fermi estimation, despite its natural suitability for aggregation: item-level biases shared across the samples of a model account for about 87% of the squared error, so averaging reduces error by only about 6%. Combining model families helps on Fermi estimation but does not surpass the strongest member on disjunctive tasks. These results show that task structure, together with the mechanism combining member outputs, is a fundamental determinant of team scaling. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-28 in cs.MA). Categorie arXiv: Artificial Intelligence (cs.AI) ; Multiagent Systems (cs.MA). Pagina del lavoro: https://arxiv.org/abs/2609.31563 — 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 2026-10-08T01:11:54.462Z dalla pagina https://arxiv.org/catchup/cs.MA/2026-09-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).