Metraipolis › archive › dc_db61878b9256
Network topology shapes consensus and vulnerability in LLM populations — arXiv:2610.31165
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.31165
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-23T15:15:21.962Z
- SHA-256 of the archived text
1541f87212f1b2bf6dae3c451e9a5593cf699dda9e2ab509ca402d3368824953- Stable URL of this page
https://metraipolis.com/documento/dc_db61878b9256
How to cite
Network topology shapes consensus and vulnerability in LLM populations — arXiv:2610.31165 — arXiv (en), https://arxiv.org/abs/2610.31165 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_db61878b9256 (Metraipolis archive dc_db61878b9256, sha256 1541f87212f1b2bf6dae3c451e9a5593cf699dda9e2ab509ca402d3368824953)
BibTeX
@misc{dcdb61878b9256,
title = {Network topology shapes consensus and vulnerability in LLM populations — arXiv:2610.31165},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.31165},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_db61878b9256; sha256 1541f87212f1b2bf6dae3c451e9a5593cf699dda9e2ab509ca402d3368824953},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.31165. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
As large language models (LLMs) are deployed in interacting populations, their collective behaviour depends not only on the models but also on who interacts with whom. Studies of convention formation in LLM populations have mainly tested well-mixed interactions. We place policies extracted from four LLMs on fully connected, random regular, and heterogeneous networks, with $10^3$ to $10^5$ agents playing the naming game. Topology affects how quickly consensus emerges, sometimes whether it emerges at all, and, in rare cases, which convention wins. Sparsest networks slow consensus up to twelvefold. Yet, by preserving correlations between persistent neighbours, they restore consensus in some cases that fail to converge under well-mixed interactions. Networks also govern vulnerability to adversarial attack. We measure the critical mass of committed agents needed to overturn an established convention. Tipping thresholds follow the same ordering across models and conditions: hubs $\leq$ random $\leq$ low-degree nodes. On the most heterogeneous topology, median critical masses are 3%, 16%, and 35%, respectively; targeting hubs requires one-sixth as many committed agents as random placement. For most models, networked populations are easier to tip than well-mixed ones, so well-mixed estimates can overstate robustness. Assessing the safety of multi-agent AI requires mapping who interacts with whom. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Physics and Society (physics.soc-ph) ; Artificial Intelligence (cs.AI); Computers and Society (cs.CY). Pagina del lavoro: https://arxiv.org/abs/2610.31165 — 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 2026-10-23T15:09:08.592Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).