AI Paper Atlas

Research map / Architectures and efficiency / Neural architecture search

HyperNAS: Enhancing Architecture Representation for NAS Predictor via Hypernetwork

Jindi Lv, Yuhao Zhou, Yuxin Tian, Qing Ye, Wentao Feng, Jiancheng Lv

CVPR 2026 · Architectures and efficiency › Neural architecture search · approach: Efficient NAS search strategies · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use global architecture encoding (cell features chained across cells) instead of isolated cell encoding (ablation)GCN predictor (NP vs HyperNAS-P) · NAS-Bench-101Kendall's TauImproved
add shared hypernetwork auxiliary task (ablation)HyperNAS-P · NAS-Bench-101Kendall's TauImproved
add predictor task to hypernetwork training (multi-task) (ablation)Hypernetwork (HyperNAS-H vs HyperNAS) · NAS-Bench-101validation accuracyImproved
use preceding cell features instead of cell position embeddings (ablation)HyperNAS-P · NAS-Bench-101Kendall's TauImproved
use cell position embeddings instead of preceding cell features (ablation)HyperNAS-P · NAS-Bench-101Kendall's TauHurt
add hypernetwork to cell-position-embedding variant (ablation)HyperNAS-P + cellPos · NAS-Bench-101Kendall's TauImproved
set loss preference coefficient q to 1.5 instead of 2 (ablation)adaptive multi-task loss · NAS-Bench-101Kendall's Tau and val accuracyImproved
use HyperNAS predictorpredictor-based NAS · NAS-Bench-201 (39 samples)Kendall's Tau: 0.494 -> 0.667 vs PINATImproved
add hypernetwork auxiliary task to reduce queries (ablation)HyperNAS vs HyperNAS-P · CIFAR-10 DARTS searchbest accuracy / queries: 97.61 (1K) vs 97.60 (0.2K queries)Mixed
hypernetwork branch increases search cost (ablation)HyperNAS vs HyperNAS-P · CIFAR-10 DARTS searchcost (G·D): 0.1 -> 0.3Hurt
use HyperNAS search vs AutoFormerViT-Base search · ImageNettop-1 accuracy: 82.4 vs 82.4No clear effect
use HyperNAS vs PINAT in DARTS space transferarchitecture search · ImageNettop-1 accuracy: 75.1 -> 75.4Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy