aiqbench analysis
The paper compares centralized, consensus-based, and independent LLM agents for simulated perimeter defense under unreliable communication. The consensus team had the lowest breach rate (15%), but a low-cost LLM given more authority over bidding raised breaches by 38 points. Both LLM tiers often mistook disguised drones for birds, and a one-sentence warning in the prompt substantially reduced that failure. For readers, the main takeaways are the architecture trade-offs and the recognition gap, though the results come from a deliberately simplified simulation and may not transfer to real deployments.
Consensus Within, Competition Between: Decentralized LLM Teams for Perimeter Defense
arXiv:2610.41044v1 Announce Type: cross Abstract: The use of large language models (LLMs) to coordinate multi-agent teams raises a fundamental architectural question: whether task allocation should be centralized in a single LLM controller or distributed across autonomous LLM agents, particularly when communication is unreliable. In a deliberately simplified perimeter-defense simulation against LLM-controlled intruders, we compared a central commander, an auction-based consensus team and independent agents under five network conditions. Breach rates were 15% for the consensus team, 32% for the commander and 29% for independent agents, and of the different network conditions, only limited radio range changed a team's breach rate, raising the commander's. In consensus team, giving a low-cost LLM more authority over the bid computation raised breaches by 38 percentage points. With birds and civilian drones in airspace, we added a photograph for each to the input text state. Only the near-frontier model treated them apart from intruders. Intruders disguised as existing taxidermy bird drones were mistook as birds by both low-cost and near-frontier LLMs, going unpursued in 59-61% of near-frontier decisions (1-16% before). However, a one-sentence forewarning in the prompt lowered it to 0-25%. We find that LLM team with a dedicated external conflict resolution protocol results in fewest breaches. Code: https://github.com/the-kas-lab/consensus-within-competition-between.
Abstract:The use of large language models (LLMs) to coordinate multi-agent teams raises a fundamental architectural question: whether task allocation should be centralized in a single LLM controller or distributed across autonomous LLM agents, particularly when communication is unreliable. In a deliberately simplified perimeter-defense simulation against LLM-controlled intruders, we compared a central commander, an auction-based consensus team and independent agents under five network conditions. Breach rates were 15% for the consensus team, 32% for the commander and 29% for independent agents, and of the different network conditions, only limited radio range changed a team's breach rate, raising the commander's. In consensus team, giving a low-cost LLM more authority over the bid computation raised breaches by 38 percentage points. With birds and civilian drones in airspace, we added a photograph for each to the input text state. Only the near-frontier model treated them apart from intruders. Intruders disguised as existing taxidermy bird drones were mistook as birds by both low-cost and near-frontier LLMs, going unpursued in 59-61% of near-frontier decisions (1-16% before). However, a one-sentence forewarning in the prompt lowered it to 0-25%. We find that LLM team with a dedicated external conflict resolution protocol results in fewest breaches. Code: this https URL.
88 pages (33 main text and references, 55 supplementary material), 7 main figures. Code: this https URL
Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Robotics (cs.RO)
arXiv:2610.41044 [cs.AI]
(or arXiv:2610.41044v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2610.41044 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 317 words
Related coverage
Decentralized Autoregressive Generation
arXiv · Jun 12, 2026
LFHE: Local-First Heuristic Evolution for Bounded Local Topology Search in Decentralized Learning with Non-IID Data
arXiv · Oct 7, 2026
SpeechLLM Meets Federated Learning for End-to-End ASR: English and Italian Case Studies
arXiv · Jul 29, 2026
Greedy Local Learning for Language Model Pretraining: Gaps and Objective Design
arXiv · Oct 7, 2026