MACA frames multi-agent coordination as posterior inference, learns a structural prior to guide orchestration, and reports 8.42% higher performance with 43.19% fewer tokens than adaptive baselines on benchmarks.
Arkil Patel, Satwik Bhattamishra, and Navin Goyal
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.MA 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Multi-Agent Coordination Adaptation via Structure-Guided Orchestration
MACA frames multi-agent coordination as posterior inference, learns a structural prior to guide orchestration, and reports 8.42% higher performance with 43.19% fewer tokens than adaptive baselines on benchmarks.