Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study investigates how state-dependent error correlations among AI agents influence voting thresholds in decision-making committees, challenging the classical assumption of independent errors. By employing a Gaussian-copula model, the authors demonstrate that shared errors can create a positive asymptotic error floor for majority voting, necessitating adjustments to the approval threshold to minimize expected loss. The findings reveal that modeling error dependence significantly improves prediction accuracy and reduces committee loss, achieving a 15.73% reduction compared to majority voting alone.
Shared errors among AI agents can drastically alter voting thresholds, leading to improved decision-making performance in committees.
The aggregation benefit of a committee of artificial intelligence (AI) agents comes from complementary information across members. Classical voting guarantees assume independent errors. Language-model errors often co-occur on the same cases. We combine Sah-Stiglitz screening with error dependence that can differ between good and bad cases. In a homogeneous exchangeable Gaussian-copula model, shared errors create a positive asymptotic error floor for majority voting and can change the approval threshold that minimizes expected loss. We estimate a heterogeneous extension from 174,384 votes cast by 28 language models on four binary-screening benchmarks. Parameters estimated from odd-indexed items predicted committee loss on even-indexed items. For the sampled committee composition, the full-matrix dependence model increased identity-line R^2 from 0.840 under independence to 0.967. In a design-balanced analysis, cost-sensitive threshold selection under independence reduced scaled loss from 60.25 for majority to 52.50. Modeling dependence reduced it further to 50.77, an incremental improvement of 1.73 units (95% bootstrap CI, 0.68-2.33). The overall reduction from majority was 15.73% (95% bootstrap CI, 13.41-16.75%).