{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/concrete-autoencoders-for-differentiable","title":"Concrete Autoencoders for Differentiable Feature Selection and Reconstruction","arxiv_id":"1901.09346","date":"2019-01-27","proceeding":null,"authors":["Abubakar Abid","Muhammad Fatih Balin","James Zou"],"abstract":"We introduce the concrete autoencoder, an end-to-end differentiable method\nfor global feature selection, which efficiently identifies a subset of the most\ninformative features and simultaneously learns a neural network to reconstruct\nthe input data from the selected features. Our method is unsupervised, and is\nbased on using a concrete selector layer as the encoder and using a standard\nneural network as the decoder. During the training phase, the temperature of\nthe concrete selector layer is gradually decreased, which encourages a\nuser-specified number of discrete features to be learned. During test time, the\nselected features can be used with the decoder network to reconstruct the\nremaining input features. We evaluate concrete autoencoders on a variety of\ndatasets, where they significantly outperform state-of-the-art methods for\nfeature selection and data reconstruction. In particular, on a large-scale gene\nexpression dataset, the concrete autoencoder selects a small subset of genes\nwhose expression levels can be use to impute the expression levels of the\nremaining genes. In doing so, it improves on the current widely-used\nexpert-curated L1000 landmark genes, potentially reducing measurement costs by\n20%. The concrete autoencoder can be implemented by adding just a few lines of\ncode to a standard autoencoder.","url_abs":"http://arxiv.org/abs/1901.09346v2","url_pdf":"http://arxiv.org/pdf/1901.09346v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"concrete-autoencoders-for-differentiable","repo_url":"https://github.com/mfbalin/Concrete-Autoencoders","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"concrete-autoencoders-for-differentiable","repo_url":"https://github.com/BarathTamas/JetBot-Thesis","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"Unlicense"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":null,"task_name":"General Classification Selection"},{"task_slug":"feature-selection","task_name":"feature selection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/general-classification-on-activity","task":"General Classification","dataset":"Activity","model":"CAE","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"42"},"uses_additional_data":false},{"leaderboard":"/sota/general-classification-on-fashion-mnist","task":"General Classification","dataset":"Fashion-MNIST","model":"CAE","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"67.7"},"uses_additional_data":false},{"leaderboard":"/sota/general-classification-on-isolet","task":"General Classification","dataset":"ISOLET","model":"CAE","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"68.5"},"uses_additional_data":false},{"leaderboard":"/sota/general-classification-on-mnist","task":"General Classification","dataset":"MNIST","model":"CAE","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"90.6"},"uses_additional_data":false},{"leaderboard":"/sota/general-classification-on-mice-protein","task":"General Classification","dataset":"Mice Protein","model":"CAE","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"13.4"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1901.09346","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1901.09346"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/BarathTamas/JetBot-Thesis","reach":{"status":"ok","spdx":"Unlicense"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/mfbalin/Concrete-Autoencoders","reach":{"status":"ok"}}],"summary":{"unverified":3},"by_repo_kind":{"listed":{"samples":3,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"baf5f98e48c1ad79","entry":"incomplete_cholesky","repo":"BarathTamas/JetBot-Thesis","repo_kind":"listed","path":"kernel_similarity/kernel_sim.py","file_url":"https://github.com/BarathTamas/JetBot-Thesis/blob/HEAD/kernel_similarity/kernel_sim.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Unlicense","inline_ok":true,"mcp_get_code":{"code_sha256":"baf5f98e48c1ad79"}},{"code_sha256_prefix":"c465ea520876eba5","entry":"incomplete_cholesky_kernel","repo":"BarathTamas/JetBot-Thesis","repo_kind":"listed","path":"kernel_similarity/kernel_sim.py","file_url":"https://github.com/BarathTamas/JetBot-Thesis/blob/HEAD/kernel_similarity/kernel_sim.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Unlicense","inline_ok":true,"mcp_get_code":{"code_sha256":"c465ea520876eba5"}},{"code_sha256_prefix":"aea989303be72a5c","entry":"vectorized_sq_dist","repo":"BarathTamas/JetBot-Thesis","repo_kind":"listed","path":"kernel_similarity/kernel_sim.py","file_url":"https://github.com/BarathTamas/JetBot-Thesis/blob/HEAD/kernel_similarity/kernel_sim.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Unlicense","inline_ok":true,"mcp_get_code":{"code_sha256":"aea989303be72a5c"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}