When Agents Disagree: The Selection Bottleneck in Multi-Agent LLM Pipelines
Applied Sciences · 14.05.2026 · 16 (10) · 4914
DOI: 10.3390/app16104914 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Sciences, online erschienen am 14.05.2026. Verfasst von Artem Maryanskyy und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Multi-agent LLM pipelines produce contradictory evidence on whether team diversity improves output quality: heterogeneous Mixture-of-Agents teams outperform single models, yet homogeneous Self-MoA teams consistently win under synthesis-based aggregation. We propose a resolution by identifying the selection bottleneck—a crossover threshold in aggregation quality that determines whether diversity helps or hurts. Under this model, we obtain a closed-form crossover threshold s∗ (Proposition 1) that separates the regimes where diversity helps and hurts. In a targeted experiment spanning 42 tasks across seven categories (N=210), a diverse team with judge-based selection achieves a win rate of 0.810 against a single-model baseline, while a homogeneous team scores 0.512—near chance (Glass’s Δ=2.07). Judge-based selection outperforms MoA-style synthesis by ΔWR=+0.631—the synthesis approach is preferred over the baseline in zero of 42 tasks by the judge panel. A decoupled evaluation with fully independent judges confirms all directional findings (Spearman ρ=0.90); win rates attenuate by 53–67% under independent evaluation relative to the primary estimates, consistent with partial measurement circularity in the judge-based cells. Exploratory evidence suggests that including a weaker model improves performance while reducing cost (p
Zitieren
APA
Maryanskyy, A., Budnikov, D., & Kaliyev, A. T. (2026). When Agents Disagree: The Selection Bottleneck in Multi-Agent LLM Pipelines. Applied Sciences, 16(10), 4914. https://doi.org/10.3390/app16104914
Vancouver
Maryanskyy A, Budnikov D, Kaliyev AT. When Agents Disagree: The Selection Bottleneck in Multi-Agent LLM Pipelines. Applied Sciences. 2026;16(10):4914. doi:10.3390/app16104914
BibTeX
@article{Maryanskyy2026,
author = {Maryanskyy, Artem and Budnikov, Dmitry and Kaliyev, Alibek T.},
title = {When Agents Disagree: The Selection Bottleneck in Multi-Agent LLM Pipelines},
journal = {Applied Sciences},
year = {2026},
volume = {16},
number = {10},
pages = {4914},
doi = {10.3390/app16104914}
}
EUVIMED