{"url":"/sota/neural-architecture-search-on-nats-bench-2","task":{"name":"Neural Architecture Search","url":"/task/architecture-search","note":null},"dataset":{"name":"NATS-Bench Topology, CIFAR-100","url":"/dataset/nats-bench"},"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":["Test Accuracy","Kendall's Tau","Spearman's Rho"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Test Accuracy":"higher","Kendall's Tau":null,"Spearman's Rho":"higher"}},"counts":{"rows":11,"rows_with_code":8,"rows_with_paper_page":11,"rows_dated":11,"rows_using_additional_data":3},"rows":[{"rank_in_archive_order":1,"model":"LayerNAS","metrics":{"Test Accuracy":"73.01±0.63"},"uses_additional_data":false,"paper_date":"2023-04-23","paper":"/paper/layernas-neural-architecture-search-in","paper_url":"https://arxiv.org/abs/2304.11517v1","paper_title":"LayerNAS: Neural Architecture Search in Polynomial Complexity","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":2,"model":"PPO (Schulman et al., 2017)","metrics":{"Test Accuracy":"71.68"},"uses_additional_data":false,"paper_date":"2017-07-20","paper":"/paper/proximal-policy-optimization-algorithms","paper_url":"http://arxiv.org/abs/1707.06347v2","paper_title":"Proximal Policy Optimization Algorithms","code":"https://github.com/labmlai/annotated_deep_learning_paper_implementations","n_code_links":188,"syntology":{"n_ran":99,"n_unverified":77,"n_samples":176,"n_pointer_only_licence":94}},{"rank_in_archive_order":3,"model":"EigenNas (Zhu et al., 2022)","metrics":{"Test Accuracy":"71.42"},"uses_additional_data":false,"paper_date":"2022-09-15","paper":"/paper/generalization-properties-of-nas-under","paper_url":"https://arxiv.org/abs/2209.07238v4","paper_title":"Generalization Properties of NAS under Activation and Skip Connection Search","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":4,"model":"RE (Real et al., 2019)","metrics":{"Test Accuracy":"71.40"},"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":5,"model":"TE-NAS (Chen et al., 2021)","metrics":{"Test Accuracy":"71.24"},"uses_additional_data":false,"paper_date":"2021-02-23","paper":"/paper/neural-architecture-search-on-imagenet-in-1","paper_url":"https://arxiv.org/abs/2102.11535v4","paper_title":"Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective","code":"https://github.com/idstcv/ZenNAS","n_code_links":4,"syntology":{"n_ran":2,"n_unverified":2,"n_samples":4,"n_pointer_only_licence":2}},{"rank_in_archive_order":6,"model":"NASI (Shu et al., 2021)","metrics":{"Test Accuracy":"71.20"},"uses_additional_data":false,"paper_date":"2021-09-02","paper":"/paper/nasi-label-and-data-agnostic-neural","paper_url":"https://arxiv.org/abs/2109.00817v2","paper_title":"NASI: Label- and Data-agnostic Neural Architecture Search at Initialization","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":7,"model":"FairNAS (Chu et al., 2021)","metrics":{"Test Accuracy":"71.00"},"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":8,"model":"KNAS (Xu et al., 2021)","metrics":{"Test Accuracy":"68.91"},"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":9,"model":"GreenMachine-1","metrics":{"Kendall's Tau":"0.705","Spearman's Rho":"0.873"},"uses_additional_data":true,"paper_date":"2024-11-22","paper":"/paper/greenmachine-automatic-design-of-zero-cost","paper_url":"https://arxiv.org/abs/2411.15290v1","paper_title":"GreenMachine: Automatic Design of Zero-Cost Proxies for Energy-Efficient NAS","code":"https://github.com/RodriguesGabriel/greenmachine","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"GreenMachine-2","metrics":{"Kendall's Tau":"0.558","Spearman's Rho":"0.741"},"uses_additional_data":true,"paper_date":"2024-11-22","paper":"/paper/greenmachine-automatic-design-of-zero-cost","paper_url":"https://arxiv.org/abs/2411.15290v1","paper_title":"GreenMachine: Automatic Design of Zero-Cost Proxies for Energy-Efficient NAS","code":"https://github.com/RodriguesGabriel/greenmachine","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"GreenMachine-3","metrics":{"Kendall's Tau":"0.296","Spearman's Rho":"0.448"},"uses_additional_data":true,"paper_date":"2024-11-22","paper":"/paper/greenmachine-automatic-design-of-zero-cost","paper_url":"https://arxiv.org/abs/2411.15290v1","paper_title":"GreenMachine: Automatic Design of Zero-Cost Proxies for Energy-Efficient NAS","code":"https://github.com/RodriguesGabriel/greenmachine","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 6,264 of the 9,581 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":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+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":3,"rows_with_any_sample_ran":3,"distinct_papers_with_graph_line":3,"distinct_papers_with_any_sample_ran":3,"samples_over_distinct_papers":{"n_ran":107,"n_unverified":80,"n_samples":187,"n_pointer_only_licence":96,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":107,"n_unverified":80,"n_samples":187,"n_pointer_only_licence":96,"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"}}}