{"url":"/sota/neural-architecture-search-on-nas-bench-201-1","task":{"name":"Neural Architecture Search","url":"/task/architecture-search","note":null},"dataset":{"name":"NAS-Bench-201, CIFAR-10","url":"/dataset/nas-bench-201"},"category":"Methodology","categories":["Methodology"],"category_note":null,"description":"**Neural architecture search (NAS)** is a technique for automating the design of artificial neural networks (ANN), a widely used model in the field of machine learning. NAS essentially takes the process of a human manually tweaking a neural network and learning what works well, and automates this task to discover more complex architectures.\r\n\r\nImage Credit : [NAS with Reinforcement Learning](https://arxiv.org/abs/1611.01578)","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["Accuracy (Test)","Accuracy (Val)","Search time (s)"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Accuracy (Test)":"higher","Accuracy (Val)":"higher","Search time (s)":"lower"}},"counts":{"rows":37,"rows_with_code":27,"rows_with_paper_page":34,"rows_dated":34,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"DiNAS","metrics":{"Accuracy (Test)":"94.37","Accuracy (Val)":"91.61","Search time (s)":"15.36"},"uses_additional_data":false,"paper_date":"2024-03-09","paper":"/paper/multi-conditioned-graph-diffusion-for-neural","paper_url":"https://arxiv.org/abs/2403.06020v2","paper_title":"Multi-conditioned Graph Diffusion for Neural Architecture Search","code":"https://github.com/rohanasthana/dinas","n_code_links":1,"syntology":{"n_ran":5,"n_unverified":2,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"Shapley-NAS","metrics":{"Accuracy (Test)":"94.37","Accuracy (Val)":"91.61"},"uses_additional_data":false,"paper_date":"2022-06-20","paper":"/paper/shapley-nas-discovering-operation-1","paper_url":"https://arxiv.org/abs/2206.09811v1","paper_title":"Shapley-NAS: Discovering Operation Contribution for Neural Architecture Search","code":"https://github.com/euphoria16/shapley-nas","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":3,"model":"AG-Net","metrics":{"Accuracy (Test)":"94.37","Accuracy (Val)":"91.61"},"uses_additional_data":false,"paper_date":"2022-03-16","paper":"/paper/learning-where-to-look-generative-nas-is","paper_url":"https://arxiv.org/abs/2203.08734v2","paper_title":"Learning Where To Look -- Generative NAS is Surprisingly Efficient","code":"https://github.com/jovitalukasik/AG-Net","n_code_links":1,"syntology":{"n_ran":8,"n_unverified":7,"n_samples":15,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"β-DARTS","metrics":{"Accuracy (Test)":"94.36","Accuracy (Val)":"91.55"},"uses_additional_data":false,"paper_date":"2022-03-03","paper":"/paper/b-darts-beta-decay-regularization-for","paper_url":"https://arxiv.org/abs/2203.01665v2","paper_title":"$β$-DARTS: Beta-Decay Regularization for Differentiable Architecture Search","code":"https://github.com/Sunshine-Ye/Beta-DARTS","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"DrNAS","metrics":{"Accuracy (Test)":"94.36","Accuracy (Val)":"91.55"},"uses_additional_data":false,"paper_date":"2020-06-18","paper":"/paper/drnas-dirichlet-neural-architecture-search","paper_url":"https://arxiv.org/abs/2006.10355v4","paper_title":"DrNAS: Dirichlet Neural Architecture Search","code":"https://github.com/xiangning-chen/DrNAS","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"Λ-DARTS","metrics":{"Accuracy (Test)":"94.36","Accuracy (Val)":"91.55"},"uses_additional_data":false,"paper_date":"2022-10-14","paper":"/paper/l-darts-mitigating-performance-collapse-by","paper_url":"https://arxiv.org/abs/2210.07998v2","paper_title":"$Λ$-DARTS: Mitigating Performance Collapse by Harmonizing Operation Selection among Cells","code":"https://github.com/dr-faustus/lambda-darts","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":7,"model":"IS-DARTS","metrics":{"Accuracy (Test)":" 94.36","Accuracy (Val)":"91.55"},"uses_additional_data":false,"paper_date":"2023-12-19","paper":"/paper/is-darts-stabilizing-darts-through-precise","paper_url":"https://arxiv.org/abs/2312.12648v1","paper_title":"IS-DARTS: Stabilizing DARTS through Precise Measurement on Candidate Importance","code":"https://github.com/hy-he/is-darts","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"CR-LSO","metrics":{"Accuracy (Test)":"94.35","Accuracy (Val)":"91.54"},"uses_additional_data":false,"paper_date":"2022-11-11","paper":"/paper/cr-lso-convex-neural-architecture","paper_url":"https://arxiv.org/abs/2211.05950v1","paper_title":"CR-LSO: Convex Neural Architecture Optimization in the Latent Space of Graph Variational Autoencoder with Input Convex Neural Networks","code":"https://github.com/raoxuan-1998/cr-lso","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"TNAS","metrics":{"Accuracy (Test)":"94.35"},"uses_additional_data":false,"paper_date":"2022-04-11","paper":"/paper/when-nas-meets-trees-an-efficient-algorithm","paper_url":"https://arxiv.org/abs/2204.04918v1","paper_title":"When NAS Meets Trees: An Efficient Algorithm for Neural Architecture Search","code":"https://github.com/guochengqian/tnas","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"BaLeNAS-TF","metrics":{"Accuracy (Test)":"94.33","Accuracy (Val)":"91.52"},"uses_additional_data":false,"paper_date":"2021-11-25","paper":"/paper/balenas-differentiable-architecture-search","paper_url":"https://arxiv.org/abs/2111.13204v1","paper_title":"BaLeNAS: Differentiable Architecture Search via the Bayesian Learning Rule","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":11,"model":"NAR","metrics":{"Accuracy (Test)":"94.33","Accuracy (Val)":"91.44"},"uses_additional_data":false,"paper_date":"2022-01-30","paper":"/paper/neural-architecture-ranker","paper_url":"https://arxiv.org/abs/2201.12725v2","paper_title":"Generalized Global Ranking-Aware Neural Architecture Ranker for Efficient Image Classifier Search","code":"https://github.com/albertipot/nar","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"α-DARTS","metrics":{"Accuracy (Test)":"94.3","Accuracy (Val)":"91.49"},"uses_additional_data":false,"paper_date":"2022-11-18","paper":"/paper/a-darts-once-more-enhancing-differentiable","paper_url":"https://arxiv.org/abs/2211.10105v1","paper_title":"$α$ DARTS Once More: Enhancing Differentiable Architecture Search by Masked Image Modeling","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":13,"model":"RF-DARTS","metrics":{"Accuracy (Test)":"94.27","Accuracy (Val)":"91.3"},"uses_additional_data":false,"paper_date":"2022-08-18","paper":"/paper/differentiable-architecture-search-with-1","paper_url":"https://arxiv.org/abs/2208.08835v1","paper_title":"Differentiable Architecture Search with Random Features","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":14,"model":"arch2vec","metrics":{"Accuracy (Test)":"94.18","Accuracy (Val)":"91.41","Search time (s)":"12000"},"uses_additional_data":false,"paper_date":"2020-06-12","paper":"/paper/does-unsupervised-architecture-representation","paper_url":"https://arxiv.org/abs/2006.06936v2","paper_title":"Does Unsupervised Architecture Representation Learning Help Neural Architecture Search?","code":"https://github.com/MSU-MLSys-Lab/arch2vec","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":9,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":15,"model":"GenNAS","metrics":{"Accuracy (Test)":"94.18","Accuracy (Val)":"-","Search time (s)":"1080"},"uses_additional_data":false,"paper_date":"2021-08-04","paper":"/paper/generic-neural-architecture-search-via","paper_url":"https://arxiv.org/abs/2108.01899v2","paper_title":"Generic Neural Architecture Search via Regression","code":"https://github.com/leeyeehoo/GenNAS","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":16,"model":"GAEA DARTS (ERM)","metrics":{"Accuracy (Test)":"94.1"},"uses_additional_data":false,"paper_date":"2020-04-16","paper":"/paper/geometry-aware-gradient-algorithms-for-neural","paper_url":"https://arxiv.org/abs/2004.07802v5","paper_title":"Geometry-Aware Gradient Algorithms for Neural Architecture Search","code":"https://github.com/liamcli/gaea_release","n_code_links":1,"syntology":{"n_ran":12,"n_unverified":3,"n_samples":15,"n_pointer_only_licence":5}},{"rank_in_archive_order":17,"model":"PRE-NAS","metrics":{"Accuracy (Test)":"94.04","Accuracy (Val)":"91.37"},"uses_additional_data":false,"paper_date":"2022-04-27","paper":"/paper/pre-nas-predictor-assisted-evolutionary","paper_url":"https://arxiv.org/abs/2204.12726v1","paper_title":"PRE-NAS: Predictor-assisted Evolutionary Neural Architecture Search","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":18,"model":"CDARTS","metrics":{"Accuracy (Test)":"94.02","Accuracy (Val)":"91.12"},"uses_additional_data":false,"paper_date":"2020-06-18","paper":"/paper/cyclic-differentiable-architecture-search","paper_url":"https://arxiv.org/abs/2006.10724v4","paper_title":"Cyclic Differentiable Architecture Search","code":"https://github.com/microsoft/cream","n_code_links":3,"syntology":null},{"rank_in_archive_order":19,"model":"GAE","metrics":{"Accuracy (Test)":" 93.99","Accuracy (Val)":"91.26"},"uses_additional_data":false,"paper_date":"2022-07-22","paper":"/paper/guided-evolutionary-neural-architecture","paper_url":"https://arxiv.org/abs/2208.06475v1","paper_title":"Guided Evolutionary Neural Architecture Search With Efficient Performance Estimation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":20,"model":"DU-DARTS","metrics":{"Accuracy (Test)":" 93.86","Accuracy (Val)":"91.21"},"uses_additional_data":false,"paper_date":"2021-11-23","paper":"/paper/du-darts-decreasing-the-uncertainty-of","paper_url":"https://www.bmvc2021-virtualconference.com/conference/papers/paper_0209.html","paper_title":"DU-DARTS: Decreasing the Uncertainty of Differentiable Architecture Search","code":"https://github.com/ShunLu91/DU-DARTS","n_code_links":1,"syntology":null},{"rank_in_archive_order":21,"model":"DARTS-","metrics":{"Accuracy (Test)":"93.80","Accuracy (Val)":"91.03","Search time (s)":"11520"},"uses_additional_data":false,"paper_date":"2020-09-02","paper":"/paper/darts-robustly-stepping-out-of-performance","paper_url":"https://arxiv.org/abs/2009.01027v2","paper_title":"DARTS-: Robustly Stepping out of Performance Collapse Without Indicators","code":"https://github.com/Meituan-AutoML/DARTS-","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":8,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":22,"model":"GDAS","metrics":{"Accuracy (Test)":"93.61","Accuracy (Val)":"89.89","Search time (s)":"28926"},"uses_additional_data":false,"paper_date":"2019-10-10","paper":"/paper/searching-for-a-robust-neural-architecture-in-1","paper_url":"https://arxiv.org/abs/1910.04465v2","paper_title":"Searching for A Robust Neural Architecture in Four GPU Hours","code":"https://github.com/D-X-Y/GDAS","n_code_links":6,"syntology":null},{"rank_in_archive_order":23,"model":"iDARTS","metrics":{"Accuracy (Test)":"93.58","Accuracy (Val)":"89.86"},"uses_additional_data":false,"paper_date":"2021-06-21","paper":"/paper/idarts-differentiable-architecture-search","paper_url":"https://arxiv.org/abs/2106.10784v1","paper_title":"iDARTS: Differentiable Architecture Search with Stochastic Implicit Gradients","code":"https://github.com/MiaoZhang0525/iDARTS","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"KNAS (k=40)","metrics":{"Accuracy (Test)":"93.43"},"uses_additional_data":false,"paper_date":"2021-11-26","paper":"/paper/knas-green-neural-architecture-search","paper_url":"https://arxiv.org/abs/2111.13293v1","paper_title":"KNAS: Green Neural Architecture Search","code":"https://github.com/jingjing-nlp/knas","n_code_links":1,"syntology":{"n_ran":6,"n_unverified":1,"n_samples":7,"n_pointer_only_licence":0}},{"rank_in_archive_order":25,"model":"FairNAS","metrics":{"Accuracy (Test)":"93.23","Accuracy (Val)":"90.07","Search time (s)":"9845"},"uses_additional_data":false,"paper_date":"2019-07-03","paper":"/paper/fairnas-rethinking-evaluation-fairness-of","paper_url":"https://arxiv.org/abs/1907.01845v5","paper_title":"FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search","code":"https://github.com/xiaomi-automl/FairNAS","n_code_links":2,"syntology":null},{"rank_in_archive_order":26,"model":"NAS-LID+RSPS","metrics":{"Accuracy (Test)":"92.9","Accuracy (Val)":"89.74"},"uses_additional_data":false,"paper_date":"2022-11-23","paper":"/paper/nas-lid-efficient-neural-architecture-search","paper_url":"https://arxiv.org/abs/2211.12759v2","paper_title":"NAS-LID: Efficient Neural Architecture Search with Local Intrinsic Dimension","code":"https://github.com/marsggbo/nas-lid","n_code_links":1,"syntology":null},{"rank_in_archive_order":27,"model":"SNAS","metrics":{"Accuracy (Test)":"92.77","Accuracy (Val)":"90.10"},"uses_additional_data":false,"paper_date":"2018-12-24","paper":"/paper/snas-stochastic-neural-architecture-search","paper_url":"https://arxiv.org/abs/1812.09926v3","paper_title":"SNAS: Stochastic Neural Architecture Search","code":"https://github.com/SNAS-Series/SNAS-Series","n_code_links":2,"syntology":null},{"rank_in_archive_order":28,"model":"EPE-NAS (N=10)","metrics":{"Accuracy (Test)":"92.63","Accuracy (Val)":"89.90","Search time (s)":"2.3"},"uses_additional_data":false,"paper_date":"2021-02-16","paper":"/paper/epe-nas-efficient-performance-estimation","paper_url":"https://arxiv.org/abs/2102.08099v1","paper_title":"EPE-NAS: Efficient Performance Estimation Without Training for Neural Architecture Search","code":"https://github.com/VascoLopes/EPENAS","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":29,"model":"EPE-NAS (N=500)","metrics":{"Accuracy (Test)":"92.27","Accuracy (Val)":"88.17","Search time (s)":"105.8"},"uses_additional_data":false,"paper_date":"2021-02-16","paper":"/paper/epe-nas-efficient-performance-estimation","paper_url":"https://arxiv.org/abs/2102.08099v1","paper_title":"EPE-NAS: Efficient Performance Estimation Without Training for Neural Architecture Search","code":"https://github.com/VascoLopes/EPENAS","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":30,"model":"EPE-NAS (N=100)","metrics":{"Accuracy (Test)":"91.59","Accuracy (Val)":"88.74","Search time (s)":"20.5"},"uses_additional_data":false,"paper_date":"2021-02-16","paper":"/paper/epe-nas-efficient-performance-estimation","paper_url":"https://arxiv.org/abs/2102.08099v1","paper_title":"EPE-NAS: Efficient Performance Estimation Without Training for Neural Architecture Search","code":"https://github.com/VascoLopes/EPENAS","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":31,"model":"EPE-NAS (N=1000)","metrics":{"Accuracy (Test)":"91.31","Accuracy (Val)":"87.87","Search time (s)":"206.2"},"uses_additional_data":false,"paper_date":"2021-02-16","paper":"/paper/epe-nas-efficient-performance-estimation","paper_url":"https://arxiv.org/abs/2102.08099v1","paper_title":"EPE-NAS: Efficient Performance Estimation Without Training for Neural Architecture Search","code":"https://github.com/VascoLopes/EPENAS","n_code_links":2,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":32,"model":"RSPS","metrics":{"Accuracy (Test)":"87.66","Accuracy (Val)":"84.16","Search time (s)":"7587"},"uses_additional_data":false,"paper_date":"2019-02-20","paper":"/paper/random-search-and-reproducibility-for-neural","paper_url":"https://arxiv.org/abs/1902.07638v3","paper_title":"Random Search and Reproducibility for Neural Architecture Search","code":"https://github.com/microsoft/nn-meter","n_code_links":4,"syntology":{"n_ran":2,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":1}},{"rank_in_archive_order":33,"model":"SETN","metrics":{"Accuracy (Test)":"86.19","Accuracy (Val)":"82.25","Search time (s)":"31010"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":34,"model":"DARTS-V1","metrics":{"Accuracy (Test)":"54.30","Accuracy (Val)":"39.77","Search time (s)":"10890"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":35,"model":"ENAS","metrics":{"Accuracy (Test)":"54.3","Accuracy (Val)":"39.77","Search time (s)":"13315"},"uses_additional_data":false,"paper_date":"2018-07-01","paper":"/paper/efficient-neural-architecture-search-via","paper_url":"https://icml.cc/Conferences/2018/Schedule?showEvent=2247","paper_title":"Efficient Neural Architecture Search via Parameters Sharing","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":36,"model":"DARTS-V2","metrics":{"Accuracy (Test)":"54.30","Accuracy (Val)":"39.77","Search time (s)":"29902"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":37,"model":"CATCH-meta","metrics":{"Accuracy (Val)":"91.33"},"uses_additional_data":false,"paper_date":"2020-07-18","paper":"/paper/catch-context-based-meta-reinforcement","paper_url":"https://arxiv.org/abs/2007.09380v3","paper_title":"CATCH: Context-based Meta Reinforcement Learning for Transferrable Architecture Search","code":null,"n_code_links":0,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,821 of the 9,623 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9623,"papers_checked":6821,"papers_extracted_not_yet_verified":65,"boards_without_verdict":29,"papers_not_yet_extracted":2737},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-25T09:33:49+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":14,"rows_with_any_sample_ran":10,"distinct_papers_with_graph_line":11,"distinct_papers_with_any_sample_ran":7,"samples_over_distinct_papers":{"n_ran":39,"n_unverified":39,"n_samples":78,"n_pointer_only_licence":13,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":48,"n_unverified":39,"n_samples":87,"n_pointer_only_licence":13,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}