AI RESEARCH
Counterfactual Graph for Multi-Agent LLM Calibration
arXiv CS.CL
•
ArXi:2605.30653v1 Announce Type: new Multi-agent LLM systems often treat agreement as evidence: when many agents in a panel give the same answer, that answer is assumed to be reliable. We show that this assumption can fail after agents communicate. Communication can induce correlated failures and false consensus, so the same vote share may reflect reliable agreement in one topology but over-confidence in another. We propose CAGE-CAL, a counterfactual agent-graph calibration framework for multi-agent LLMs.