Metraipolis › archive › dc_c0dededd9896
Theory of Scene: Breaking the Symmetry Trap in Multi-Agent LLM Coordination — arXiv:2609.32939
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.32939
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T20:54:08.348Z
- SHA-256 of the archived text
5fbf4fce4021220840240e22552e10cf66c4566b2872e75ef4c3ad6f4fdf230c- Stable URL of this page
https://metraipolis.com/documento/dc_c0dededd9896
How to cite
Theory of Scene: Breaking the Symmetry Trap in Multi-Agent LLM Coordination — arXiv:2609.32939 — arXiv (en), https://arxiv.org/abs/2609.32939 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c0dededd9896 (Metraipolis archive dc_c0dededd9896, sha256 5fbf4fce4021220840240e22552e10cf66c4566b2872e75ef4c3ad6f4fdf230c)
BibTeX
@misc{dcc0dededd9896,
title = {Theory of Scene: Breaking the Symmetry Trap in Multi-Agent LLM Coordination — arXiv:2609.32939},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.32939},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c0dededd9896; sha256 5fbf4fce4021220840240e22552e10cf66c4566b2872e75ef4c3ad6f4fdf230c},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.32939. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Multi-agent systems built on large language models (LLMs) are largely homogeneous, as their agents behave alike even across distinct LLMs. We show that when such agents act concurrently without communication, they collide on targets they must split and diverge on targets they must take together, a double failure we term the symmetry trap. Theory of Mind (ToM), widely used for coordination without communication, cannot escape this trap, since homogeneous agents form the same prediction of one another and respond to it in the same way. We propose Theory of Scene (ToS), a training-free reasoning schema in which each agent reads its public role, the only difference between the agents, and the task context they all observe. Homogeneous agents thereby derive one division of labor, each taking the part its role fixes, which turns homogeneity from the cause of the trap into the cure. ToS reads the role together with the scene through role gating, which determines whether ownership overlaps or is already divided, and the task context through task coupling, which infers whether the team must converge on each target, divide it, or take its stages in turn. We evaluate on DivvyBench, a controlled environment we introduce, whose target types make an episode Competitive, Cooperative, or Mixed across Tabletop, Airspace, and Household scenarios, and on two established agentic benchmarks, GovSim and Overcooked. ToS outperforms all six baselines on every benchmark, and each baseline falls far behind it in at least one setting. Against ToM given the same inputs, ToS raises the DivvyBench success rate from 71.1% to 99.6%, the GovSim total gain from 207 to 400, and the Overcooked level-normalized throughput from 1.41 to 1.67. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.MA). Categorie arXiv: Machine Learning (cs.LG) ; Computation and Language (cs.CL); Multiagent Systems (cs.MA). Pagina del lavoro: https://arxiv.org/abs/2609.32939 — 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 2026-10-07T20:24:41.684Z dalla pagina https://arxiv.org/catchup/cs.MA/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).