{"url":"/sota/sequential-image-classification-on-sequential","task":{"name":"Sequential Image Classification","url":"/task/sequential-image-classification","note":null},"dataset":{"name":"Sequential MNIST","url":"/dataset/mnist"},"category":"Computer Vision","categories":["Adversarial","Computer Vision"],"category_note":null,"description":"Sequential image classification is the task of classifying a sequence of images.\r\n\r\n<span style=\"color:grey; opacity: 0.6\">( Image credit: [TensorFlow-101](https://github.com/sjchoi86/tensorflow-101/blob/master/notebooks/rnn_mnist_simple.ipynb) )</span>","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":["Permuted Accuracy","Unpermuted Accuracy"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Permuted Accuracy":"higher","Unpermuted Accuracy":"higher"}},"counts":{"rows":30,"rows_with_code":30,"rows_with_paper_page":30,"rows_dated":30,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"SMPConv","metrics":{"Permuted Accuracy":"99.10","Unpermuted Accuracy":"99.75"},"uses_additional_data":false,"paper_date":"2023-04-05","paper":"/paper/smpconv-self-moving-point-representations-for","paper_url":"https://arxiv.org/abs/2304.02330v1","paper_title":"SMPConv: Self-moving Point Representations for Continuous Convolution","code":"https://github.com/sangnekim/smpconv","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":7,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"LSSL","metrics":{"Permuted Accuracy":"98.76%","Unpermuted Accuracy":"99.53%"},"uses_additional_data":false,"paper_date":"2021-10-26","paper":"/paper/combining-recurrent-convolutional-and","paper_url":"https://arxiv.org/abs/2110.13985v1","paper_title":"Combining Recurrent, Convolutional, and Continuous-time Models with Linear State-Space Layers","code":"https://github.com/hazyresearch/state-spaces","n_code_links":2,"syntology":{"n_ran":19,"n_unverified":34,"n_samples":53,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"FlexTCN-4","metrics":{"Permuted Accuracy":"98.72%"},"uses_additional_data":false,"paper_date":"2021-10-15","paper":"/paper/flexconv-continuous-kernel-convolutions-with-1","paper_url":"https://arxiv.org/abs/2110.08059v3","paper_title":"FlexConv: Continuous Kernel Convolutions with Differentiable Kernel Sizes","code":"https://github.com/rjbruin/flexconv","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"S4","metrics":{"Permuted Accuracy":"98.70%","Unpermuted Accuracy":"99.63%"},"uses_additional_data":false,"paper_date":"2021-10-31","paper":"/paper/efficiently-modeling-long-sequences-with-1","paper_url":"https://arxiv.org/abs/2111.00396v3","paper_title":"Efficiently Modeling Long Sequences with Structured State Spaces","code":"https://github.com/state-spaces/s4","n_code_links":8,"syntology":{"n_ran":28,"n_unverified":27,"n_samples":55,"n_pointer_only_licence":3}},{"rank_in_archive_order":5,"model":"CKCNN (1M)","metrics":{"Permuted Accuracy":"98.54%","Unpermuted Accuracy":"99.32%"},"uses_additional_data":false,"paper_date":"2021-02-04","paper":"/paper/ckconv-continuous-kernel-convolution-for","paper_url":"https://arxiv.org/abs/2102.02611v3","paper_title":"CKConv: Continuous Kernel Convolution For Sequential Data","code":"https://github.com/dwromero/ckconv","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"Modified LMU (165k)","metrics":{"Permuted Accuracy":"98.49%"},"uses_additional_data":false,"paper_date":"2021-02-22","paper":"/paper/parallelizing-legendre-memory-unit-training","paper_url":"https://arxiv.org/abs/2102.11417v2","paper_title":"Parallelizing Legendre Memory Unit Training","code":"https://github.com/hrshtv/pytorch-lmu","n_code_links":2,"syntology":null},{"rank_in_archive_order":7,"model":"UnICORNN","metrics":{"Permuted Accuracy":"98.4"},"uses_additional_data":false,"paper_date":"2021-03-09","paper":"/paper/unicornn-a-recurrent-model-for-learning-very","paper_url":"https://arxiv.org/abs/2103.05487v2","paper_title":"UnICORNN: A recurrent model for learning very long time dependencies","code":"https://github.com/tk-rusch/unicornn","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":8,"model":"HiPPO-LegS","metrics":{"Permuted Accuracy":"98.3%"},"uses_additional_data":false,"paper_date":"2020-08-17","paper":"/paper/hippo-recurrent-memory-with-optimal","paper_url":"https://arxiv.org/abs/2008.07669v2","paper_title":"HiPPO: Recurrent Memory with Optimal Polynomial Projections","code":"https://github.com/HazyResearch/hippo-code","n_code_links":2,"syntology":{"n_ran":6,"n_unverified":6,"n_samples":12,"n_pointer_only_licence":2}},{"rank_in_archive_order":9,"model":"CKCNN (100k)","metrics":{"Permuted Accuracy":"98%","Unpermuted Accuracy":"99.31%"},"uses_additional_data":false,"paper_date":"2021-02-04","paper":"/paper/ckconv-continuous-kernel-convolution-for","paper_url":"https://arxiv.org/abs/2102.02611v3","paper_title":"CKConv: Continuous Kernel Convolution For Sequential Data","code":"https://github.com/dwromero/ckconv","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"ODE-LSTM","metrics":{"Permuted Accuracy":"97.83%"},"uses_additional_data":false,"paper_date":"2020-06-08","paper":"/paper/learning-long-term-dependencies-in","paper_url":"https://arxiv.org/abs/2006.04418v4","paper_title":"Learning Long-Term Dependencies in Irregularly-Sampled Time Series","code":"https://github.com/mlech26l/learning-long-term-irregular-ts","n_code_links":2,"syntology":null},{"rank_in_archive_order":11,"model":"coRNN","metrics":{"Permuted Accuracy":"97.34%","Unpermuted Accuracy":"99.4%"},"uses_additional_data":false,"paper_date":"2020-10-02","paper":"/paper/coupled-oscillatory-recurrent-neural-network","paper_url":"https://arxiv.org/abs/2010.00951v2","paper_title":"Coupled Oscillatory Recurrent Neural Network (coRNN): An accurate and (gradient) stable architecture for learning long time dependencies","code":"https://github.com/tk-rusch/coRNN","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":2,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":12,"model":"Dense IndRNN","metrics":{"Permuted Accuracy":"97.2%","Unpermuted Accuracy":"99.48%"},"uses_additional_data":false,"paper_date":"2019-10-11","paper":"/paper/deep-independently-recurrent-neural-network","paper_url":"https://arxiv.org/abs/1910.06251v3","paper_title":"Deep Independently Recurrent Neural Network (IndRNN)","code":"https://github.com/Sunnydreamrain/IndRNN_pytorch","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"Temporal Convolutional Network","metrics":{"Permuted Accuracy":"97.2%","Unpermuted Accuracy":"99.0%"},"uses_additional_data":false,"paper_date":"2018-03-04","paper":"/paper/an-empirical-evaluation-of-generic","paper_url":"http://arxiv.org/abs/1803.01271v2","paper_title":"An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling","code":"https://github.com/timeseriesAI/tsai/tree/main/tsai/models","n_code_links":35,"syntology":{"n_ran":2,"n_unverified":8,"n_samples":10,"n_pointer_only_licence":1}},{"rank_in_archive_order":14,"model":"LMU","metrics":{"Permuted Accuracy":"97.2%"},"uses_additional_data":false,"paper_date":"2019-12-01","paper":"/paper/legendre-memory-units-continuous-time","paper_url":"http://papers.nips.cc/paper/9689-legendre-memory-units-continuous-time-representation-in-recurrent-neural-networks","paper_title":"Legendre Memory Units: Continuous-Time Representation in Recurrent Neural Networks","code":"https://github.com/abr/neurips2019","n_code_links":2,"syntology":null},{"rank_in_archive_order":15,"model":"Adaptive-saturated RNN","metrics":{"Permuted Accuracy":"96.96%","Unpermuted Accuracy":"99.3%"},"uses_additional_data":false,"paper_date":"2023-04-24","paper":"/paper/adaptive-saturated-rnn-remember-more-with-1","paper_url":"https://arxiv.org/abs/2304.11790v1","paper_title":"Adaptive-saturated RNN: Remember more with less instability","code":"https://github.com/ndminhkhoi46/asRNN","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":3}},{"rank_in_archive_order":16,"model":"Sparse Combo Net","metrics":{"Permuted Accuracy":"96.94"},"uses_additional_data":false,"paper_date":"2021-06-16","paper":"/paper/recursive-construction-of-stable-assemblies","paper_url":"https://arxiv.org/abs/2106.08928v6","paper_title":"RNNs of RNNs: Recursive Construction of Stable Assemblies of Recurrent Neural Networks","code":"https://github.com/ennisthemennis/sparse-combo-net","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"GAM-RHN-1","metrics":{"Permuted Accuracy":"96.8%"},"uses_additional_data":false,"paper_date":"2019-12-13","paper":"/paper/recurrent-highway-networks-with-grouped","paper_url":"https://ieeexplore.ieee.org/document/8932404","paper_title":"Recurrent Highway Networks with Grouped Auxiliary Memory","code":"https://github.com/WilliamRo/gam_rhn","n_code_links":4,"syntology":null},{"rank_in_archive_order":18,"model":"LEM","metrics":{"Permuted Accuracy":"96.6%","Unpermuted Accuracy":"99.5%"},"uses_additional_data":false,"paper_date":"2021-10-10","paper":"/paper/long-expressive-memory-for-sequence-modeling-1","paper_url":"https://arxiv.org/abs/2110.04744v2","paper_title":"Long Expressive Memory for Sequence Modeling","code":"https://github.com/tk-rusch/lem","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":9,"n_samples":10,"n_pointer_only_licence":1}},{"rank_in_archive_order":19,"model":"LipschitzRNN","metrics":{"Permuted Accuracy":"96.3%","Unpermuted Accuracy":"99.4"},"uses_additional_data":false,"paper_date":"2020-06-22","paper":"/paper/lipschitz-recurrent-neural-networks","paper_url":"https://arxiv.org/abs/2006.12070v3","paper_title":"Lipschitz Recurrent Neural Networks","code":"https://github.com/erichson/LipschitzRNN","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":20,"model":"DNC+CUW","metrics":{"Permuted Accuracy":"96.3%","Unpermuted Accuracy":"99.1%"},"uses_additional_data":false,"paper_date":"2019-01-05","paper":"/paper/learning-to-remember-more-with-less","paper_url":"http://arxiv.org/abs/1901.01347v2","paper_title":"Learning to Remember More with Less Memorization","code":"https://github.com/thaihungle/UW-DNC","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":21,"model":"IndRNN","metrics":{"Permuted Accuracy":"96%","Unpermuted Accuracy":"99%"},"uses_additional_data":false,"paper_date":"2018-03-13","paper":"/paper/independently-recurrent-neural-network-indrnn","paper_url":"http://arxiv.org/abs/1803.04831v3","paper_title":"Independently Recurrent Neural Network (IndRNN): Building A Longer and Deeper RNN","code":"https://github.com/TobiasLee/Text-Classification","n_code_links":11,"syntology":{"n_ran":1,"n_unverified":4,"n_samples":5,"n_pointer_only_licence":1}},{"rank_in_archive_order":22,"model":"BN LSTM","metrics":{"Permuted Accuracy":"95.4%","Unpermuted Accuracy":"99%"},"uses_additional_data":false,"paper_date":"2016-03-30","paper":"/paper/recurrent-batch-normalization","paper_url":"http://arxiv.org/abs/1603.09025v5","paper_title":"Recurrent Batch Normalization","code":"https://github.com/cooijmanstim/recurrent-batch-normalization","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":2,"n_samples":2,"n_pointer_only_licence":0}},{"rank_in_archive_order":23,"model":"EGRU","metrics":{"Permuted Accuracy":"95.1%","Unpermuted Accuracy":"98.3%"},"uses_additional_data":false,"paper_date":"2022-06-13","paper":"/paper/egru-event-based-gru-for-activity-sparse","paper_url":"https://arxiv.org/abs/2206.06178v3","paper_title":"Efficient recurrent architectures through activity sparsity and sparse back-propagation through time","code":"https://github.com/khaleelkhan/evnn","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"Dilated GRU","metrics":{"Permuted Accuracy":"94.6%","Unpermuted Accuracy":"99.2%"},"uses_additional_data":false,"paper_date":"2017-10-05","paper":"/paper/dilated-recurrent-neural-networks","paper_url":"http://arxiv.org/abs/1710.02224v3","paper_title":"Dilated Recurrent Neural Networks","code":"https://github.com/code-terminator/DilatedRNN","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":6,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":25,"model":"Full-capacity uRNN","metrics":{"Permuted Accuracy":"94.1%","Unpermuted Accuracy":"96.9%"},"uses_additional_data":false,"paper_date":"2016-10-31","paper":"/paper/full-capacity-unitary-recurrent-neural","paper_url":"http://arxiv.org/abs/1611.00035v1","paper_title":"Full-Capacity Unitary Recurrent Neural Networks","code":"https://github.com/stwisdom/urnn","n_code_links":2,"syntology":null},{"rank_in_archive_order":26,"model":"LSTM","metrics":{"Permuted Accuracy":"88%","Unpermuted Accuracy":"98.2%"},"uses_additional_data":false,"paper_date":"2015-11-20","paper":"/paper/unitary-evolution-recurrent-neural-networks","paper_url":"http://arxiv.org/abs/1511.06464v4","paper_title":"Unitary Evolution Recurrent Neural Networks","code":"https://github.com/v0lta/Complex-gated-recurrent-neural-networks","n_code_links":2,"syntology":{"n_ran":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1}},{"rank_in_archive_order":27,"model":"iRNN","metrics":{"Permuted Accuracy":"82%","Unpermuted Accuracy":"97%"},"uses_additional_data":false,"paper_date":"2015-04-03","paper":"/paper/a-simple-way-to-initialize-recurrent-networks","paper_url":"http://arxiv.org/abs/1504.00941v2","paper_title":"A Simple Way to Initialize Recurrent Networks of Rectified Linear Units","code":"https://github.com/facebookresearch/salina","n_code_links":6,"syntology":null},{"rank_in_archive_order":28,"model":"FlexTCN-6","metrics":{"Unpermuted Accuracy":"99.62%"},"uses_additional_data":false,"paper_date":"2021-10-15","paper":"/paper/flexconv-continuous-kernel-convolutions-with-1","paper_url":"https://arxiv.org/abs/2110.08059v3","paper_title":"FlexConv: Continuous Kernel Convolutions with Differentiable Kernel Sizes","code":"https://github.com/rjbruin/flexconv","n_code_links":1,"syntology":null},{"rank_in_archive_order":29,"model":"STAR","metrics":{"Unpermuted Accuracy":"99.4%"},"uses_additional_data":false,"paper_date":"2019-11-25","paper":"/paper/gating-revisited-deep-multi-layer-rnns-that-1","paper_url":"https://arxiv.org/abs/1911.11033v4","paper_title":"Gating Revisited: Deep Multi-layer RNNs That Can Be Trained","code":"https://github.com/0zgur0/STAR_Network","n_code_links":3,"syntology":null},{"rank_in_archive_order":30,"model":"R-Transformer","metrics":{"Unpermuted Accuracy":"99.1%"},"uses_additional_data":false,"paper_date":"2019-07-12","paper":"/paper/r-transformer-recurrent-neural-network","paper_url":"https://arxiv.org/abs/1907.05572v1","paper_title":"R-Transformer: Recurrent Neural Network Enhanced Transformer","code":"https://github.com/DSE-MSU/R-transformer","n_code_links":2,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":2}}],"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":16,"rows_with_any_sample_ran":12,"distinct_papers_with_graph_line":16,"distinct_papers_with_any_sample_ran":12,"samples_over_distinct_papers":{"n_ran":67,"n_unverified":109,"n_samples":176,"n_pointer_only_licence":16,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":67,"n_unverified":109,"n_samples":176,"n_pointer_only_licence":16,"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"}}}