Reduces the computational cost of Neural Architecture Search for ensembles from O(M) to O(1).
March 23, 2026
Original Paper
AgenticRS-EnsNAS: Ensemble-Decoupled Self-Evolving Architecture Search
arXiv · 2603.20014
The Takeaway
Standard industrial deployments use ensembles of 50-200 models, making traditional NAS prohibitively expensive. This framework uses ensemble theory to predict system-level performance from single-learner evaluation, drastically accelerating iteration cycles.
From the abstract
Neural Architecture Search (NAS) deployment in industrial production systems faces a fundamental validation bottleneck: verifying a single candidate architecture pi requires evaluating the deployed ensemble of M models, incurring prohibitive O(M) computational cost per candidate. This cost barrier severely limits architecture iteration frequency in real-world applications where ensembles (M=50-200) are standard for robustness. This work introduces Ensemble-Decoupled Architecture Search, a framew