{"url":"/sota/architecture-search-on-cifar-10-image","task":{"name":"Neural Architecture Search","url":"/task/architecture-search","note":null},"dataset":{"name":"CIFAR-10 Image Classification","url":"/dataset/cifar-10"},"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":["Percentage error","Params","FLOPS","Search Time (GPU days)"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Percentage error":"lower","Params":"lower","FLOPS":"lower","Search Time (GPU days)":"lower"}},"counts":{"rows":19,"rows_with_code":19,"rows_with_paper_page":19,"rows_dated":19,"rows_using_additional_data":6},"rows":[{"rank_in_archive_order":1,"model":"NAT-M4","metrics":{"FLOPS":"468M","Params":"6.9M","Percentage error":"1.6"},"uses_additional_data":true,"paper_date":"2020-05-12","paper":"/paper/neural-architecture-transfer","paper_url":"https://arxiv.org/abs/2005.05859v2","paper_title":"Neural Architecture Transfer","code":"https://github.com/human-analysis/neural-architecture-transfer","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":2,"model":"NAT-M3","metrics":{"FLOPS":"392M","Params":"6.2M","Percentage error":"1.8"},"uses_additional_data":true,"paper_date":"2020-05-12","paper":"/paper/neural-architecture-transfer","paper_url":"https://arxiv.org/abs/2005.05859v2","paper_title":"Neural Architecture Transfer","code":"https://github.com/human-analysis/neural-architecture-transfer","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":3,"model":"NAONet + c/o","metrics":{"Params":"144.6M","Percentage error":"1.93"},"uses_additional_data":false,"paper_date":"2018-08-22","paper":"/paper/neural-architecture-optimization","paper_url":"https://arxiv.org/abs/1808.07233v5","paper_title":"Neural Architecture Optimization","code":"https://github.com/renqianluo/NAO","n_code_links":5,"syntology":null},{"rank_in_archive_order":4,"model":"SharpSepConvDARTS","metrics":{"FLOPS":"579M","Params":"3.6M","Percentage error":"1.98"},"uses_additional_data":false,"paper_date":"2019-03-23","paper":"/paper/sharpdarts-faster-and-more-accurate","paper_url":"http://arxiv.org/abs/1903.09900v1","paper_title":"sharpDARTS: Faster and More Accurate Differentiable Architecture Search","code":"https://github.com/ahundt/sharpDARTS","n_code_links":2,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"MUXNet-m","metrics":{"FLOPS":"200M","Params":"2.1M","Percentage error":"2.0"},"uses_additional_data":true,"paper_date":"2020-03-31","paper":"/paper/muxconv-information-multiplexing-in","paper_url":"https://arxiv.org/abs/2003.13880v2","paper_title":"MUXConv: Information Multiplexing in Convolutional Neural Networks","code":"https://github.com/human-analysis/MUXConv","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":2,"n_samples":5,"n_pointer_only_licence":5}},{"rank_in_archive_order":6,"model":"Proxyless-G + c/o","metrics":{"Params":"5.7M","Percentage error":"2.08"},"uses_additional_data":true,"paper_date":"2018-12-02","paper":"/paper/proxylessnas-direct-neural-architecture","paper_url":"http://arxiv.org/abs/1812.00332v2","paper_title":"ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware","code":"https://github.com/osmr/imgclsmob","n_code_links":23,"syntology":{"n_ran":9,"n_unverified":18,"n_samples":27,"n_pointer_only_licence":0}},{"rank_in_archive_order":7,"model":"NAT-M2","metrics":{"FLOPS":"291M","Params":"4.6M","Percentage error":"2.1"},"uses_additional_data":true,"paper_date":"2020-05-12","paper":"/paper/neural-architecture-transfer","paper_url":"https://arxiv.org/abs/2005.05859v2","paper_title":"Neural Architecture Transfer","code":"https://github.com/human-analysis/neural-architecture-transfer","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":8,"model":"AmoebaNet-B + c/o","metrics":{"Params":"34.9M","Percentage error":"2.13"},"uses_additional_data":false,"paper_date":"2018-02-05","paper":"/paper/regularized-evolution-for-image-classifier","paper_url":"http://arxiv.org/abs/1802.01548v7","paper_title":"Regularized Evolution for Image Classifier Architecture Search","code":"https://github.com/tensorflow/tpu/tree/master/models/official/amoeba_net","n_code_links":5,"syntology":null},{"rank_in_archive_order":9,"model":"PathLevel EAS + c/o","metrics":{"Params":"14.3M","Percentage error":"2.30"},"uses_additional_data":false,"paper_date":"2018-06-07","paper":"/paper/path-level-network-transformation-for","paper_url":"http://arxiv.org/abs/1806.02639v1","paper_title":"Path-Level Network Transformation for Efficient Architecture Search","code":"https://github.com/han-cai/RL4AS_NetTrans","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":10,"model":"NAS-RL-A + c/o","metrics":{"Params":"27.6M","Percentage error":"2.4"},"uses_additional_data":false,"paper_date":"2016-11-05","paper":"/paper/neural-architecture-search-with-reinforcement","paper_url":"http://arxiv.org/abs/1611.01578v2","paper_title":"Neural Architecture Search with Reinforcement Learning","code":"https://github.com/tensorflow/models","n_code_links":12,"syntology":{"n_ran":5,"n_unverified":0,"n_samples":5,"n_pointer_only_licence":3}},{"rank_in_archive_order":11,"model":"CATE","metrics":{"Params":"4.1","Percentage error":"2.46","Search Time (GPU days)":"10.3"},"uses_additional_data":false,"paper_date":"2021-02-14","paper":"/paper/cate-computation-aware-neural-architecture","paper_url":"https://arxiv.org/abs/2102.07108v2","paper_title":"CATE: Computation-aware Neural Architecture Encoding with Transformers","code":"https://github.com/MSU-MLSys-Lab/CATE","n_code_links":1,"syntology":{"n_ran":9,"n_unverified":1,"n_samples":10,"n_pointer_only_licence":0}},{"rank_in_archive_order":12,"model":"Soft Parameter Sharing","metrics":{"Params":"33.5","Percentage error":"2.53"},"uses_additional_data":false,"paper_date":"2019-02-26","paper":"/paper/learning-implicitly-recurrent-cnns-through","paper_url":"http://arxiv.org/abs/1902.09701v2","paper_title":"Learning Implicitly Recurrent CNNs Through Parameter Sharing","code":"https://github.com/lolemacs/soft-sharing","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"arch2vec","metrics":{"Params":"3.6","Percentage error":"2.56","Search Time (GPU days)":"10.5"},"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":14,"model":"GATES + c/o","metrics":{"Params":"4.1M","Percentage error":"2.58"},"uses_additional_data":false,"paper_date":"2020-04-04","paper":"/paper/a-generic-graph-based-neural-architecture","paper_url":"https://arxiv.org/abs/2004.01899v3","paper_title":"A Generic Graph-based Neural Architecture Encoding Scheme for Predictor-based NAS","code":"https://github.com/walkerning/aw_nas","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":15,"model":"NAT-M1","metrics":{"FLOPS":"232M","Params":"4.3M","Percentage error":"2.6"},"uses_additional_data":true,"paper_date":"2020-05-12","paper":"/paper/neural-architecture-transfer","paper_url":"https://arxiv.org/abs/2005.05859v2","paper_title":"Neural Architecture Transfer","code":"https://github.com/human-analysis/neural-architecture-transfer","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":16,"model":"DARTS + c/o","metrics":{"Params":"3.4M","Percentage error":"2.83","Search Time (GPU days)":"4"},"uses_additional_data":false,"paper_date":"2018-06-24","paper":"/paper/darts-differentiable-architecture-search","paper_url":"http://arxiv.org/abs/1806.09055v2","paper_title":"DARTS: Differentiable Architecture Search","code":"https://github.com/google-research/google-research/tree/master/enas_lm","n_code_links":59,"syntology":{"n_ran":69,"n_unverified":87,"n_samples":156,"n_pointer_only_licence":48}},{"rank_in_archive_order":17,"model":"ENAS + c/o","metrics":{"Params":"4.6M","Percentage error":"2.89"},"uses_additional_data":false,"paper_date":"2018-02-09","paper":"/paper/efficient-neural-architecture-search-via-1","paper_url":"http://arxiv.org/abs/1802.03268v2","paper_title":"Efficient Neural Architecture Search via Parameter Sharing","code":"https://github.com/google-research/google-research/tree/master/enas_lm","n_code_links":28,"syntology":{"n_ran":18,"n_unverified":20,"n_samples":38,"n_pointer_only_licence":25}},{"rank_in_archive_order":18,"model":"EEEA-Net-C (b=5)+ CO","metrics":{"Params":"3.6"},"uses_additional_data":false,"paper_date":"2021-08-13","paper":"/paper/eeea-net-an-early-exit-evolutionary-neural","paper_url":"https://arxiv.org/abs/2108.06156v1","paper_title":"EEEA-Net: An Early Exit Evolutionary Neural Architecture Search","code":"https://github.com/chakkritte/eeea-net","n_code_links":1,"syntology":null},{"rank_in_archive_order":19,"model":"AlphaX-1 (cutout NASNet)","metrics":{"Params":"3.59M"},"uses_additional_data":false,"paper_date":"2019-03-26","paper":"/paper/alphax-exploring-neural-architectures-with-1","paper_url":"https://arxiv.org/abs/1903.11059v2","paper_title":"AlphaX: eXploring Neural Architectures with Deep Neural Networks and Monte Carlo Tree Search","code":"https://github.com/linnanwang/AlphaX-NASBench101","n_code_links":1,"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 7,081 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":7081,"papers_extracted_not_yet_verified":217,"boards_without_verdict":27,"papers_not_yet_extracted":2325},"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":8,"distinct_papers_with_graph_line":11,"distinct_papers_with_any_sample_ran":8,"samples_over_distinct_papers":{"n_ran":116,"n_unverified":140,"n_samples":256,"n_pointer_only_licence":82,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":116,"n_unverified":143,"n_samples":259,"n_pointer_only_licence":85,"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"}}}