Loading
Life sciences · Preprint
arXiv · September 10, 2026
Early or partial results. Treat as a signal, not a conclusion.
CoRA-NAS is a two-stage architecture search method combining zero-cost proxies with low-cost learning-curve refinement, reported in an unreviewed preprint. It achieves stated Spearman correlations of 0.946–0.894 across three benchmarks and 0.715 on NAS-Bench-101, and selects an architecture on CIFAR-100 with 73.32% accuracy versus a reported ground truth of 73.37%. The work is methodologically interesting but lacks peer review and real-world validation.
Benchmark evaluation study. Neural architectures sampled from four standardized benchmark search spaces (NAS-Bench-201, NAS-Bench-101, TransNAS-Bench-101, NATS-SSS).. Intervention: CoRA-NAS two-stage framework: coarse ranking via consensus proxies followed by residual refinement via learning-curve extrapolation.. Compared with: Compared methods (specific names not listed in abstract); evaluated against benchmark ground-truth architectures and accuracy labels..
Mean Spearman correlation on NAS-Bench-201: 0.946 Mean Spearman correlation on NAS-Bench-101: 0.715 Mean Spearman correlation on TransNAS-Bench-101: 0.786
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
The source did not state who this applies to in practice.
Unreviewed methodological paper introducing a neural architecture search framework with benchmark correlations but no clinical or real-world validation; claims require peer review.
As stated by the source record.
Quoted from the source exactly as published.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
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.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.