CoRA-NAS: Coarse Ranking and Anchor-Residual Refinement for Neural Architecture Search

Researchers introduced CoRA-NAS, a two-stage neural architecture search framework that combines a static ranking prior with low-cost learning-curve refinement, according to a new arXiv paper (arXiv:2609.11884v1). CoRA-Refine achieved mean Spearman correlations of 0.946 on NAS-Bench-201, 0.715 on NAS-Bench-101, 0.786 on TransNAS-Bench-101, and 0.894 on NATS-SSS, with its worst-space correlation of 0.715 the highest among compared methods. On NAS-Bench-201/CIFAR-100, the selected architecture reached 73.32% accuracy, near the reported ground-truth best of 73.37%, while using roughly 1% of the cost of fully training the candidate set.

arXiv:2609.11884v1 Announce Type: new Abstract: Zero-cost proxies rank architectures cheaply, but their reliability varies across search spaces. We introduce CoRA-NAS COarse Ranking + Anchor-residual , a two-stage framework combining a static ranking prior with low-cost learning-curve refinement. CoRA-Rank aggregates capacity and structure-at-initialization proxies through an equal-weight log-rank consensus and a target-free consensus gate. CoRA-Refine samples anchors across this prior, extrapolates their early validation curves, and propagates a learned residual correction with an ExtraTrees model. The refinement uses approximately 1% of the cost of fully training the candidate set. Fully trained architecture-accuracy labels are not used to fit the ranker. One configuration is used across spaces, with space-specific architecture encodings. Across NAS-Bench-201, NAS-Bench-101, TransNAS-Bench-101, and NATS-SSS, CoRA-Refine achieves mean Spearman correlations of 0.946, 0.715, 0.786, and 0.894, respectively. Its worst-space correlation of 0.715 is the highest among the compared methods. On NAS-Bench-201/CIFAR-100, its selected architecture reaches 73.32% accuracy, near the reported ground-truth best of 73.37%. On the pure size space, refinement recovers the static prior's shortfall relative to parameter count, while remaining tied with the strongest capacity proxies within noise. The resulting framework combines cross-space ranking robustness with low-cost architecture selection.