{"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/lossy-compression-for-lossless-prediction","title":"Lossy Compression for Lossless Prediction","arxiv_id":"2106.10800","date":"2021-06-21","proceeding":"NeurIPS 2021 12","authors":["Yann Dubois","Benjamin Bloem-Reddy","Karen Ullrich","Chris J. Maddison"],"abstract":"Most data is automatically collected and only ever \"seen\" by algorithms. Yet, data compressors preserve perceptual fidelity rather than just the information needed by algorithms performing downstream tasks. In this paper, we characterize the bit-rate required to ensure high performance on all predictive tasks that are invariant under a set of transformations, such as data augmentations. Based on our theory, we design unsupervised objectives for training neural compressors. Using these objectives, we train a generic image compressor that achieves substantial rate savings (more than $1000\\times$ on ImageNet) compared to JPEG on 8 datasets, without decreasing downstream classification performance.","url_abs":"https://arxiv.org/abs/2106.10800v5","url_pdf":"https://arxiv.org/pdf/2106.10800v5.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":"lossy-compression-for-lossless-prediction","repo_url":"https://github.com/YannDubs/lossyless","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"feature-compression","task_name":"Feature Compression"},{"task_slug":"image-compression","task_name":"Image Compression"},{"task_slug":"prediction","task_name":"Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-compression-on-cifar-10","task":"Image Compression","dataset":"CIFAR-10","model":"Lossyless Compressor","rank_in_archive_order":1,"of":1,"metrics":{"Bit rate":"1410"},"uses_additional_data":true},{"leaderboard":"/sota/image-compression-on-caltech101","task":"Image Compression","dataset":"Caltech101","model":"Lossyless Compressor","rank_in_archive_order":1,"of":1,"metrics":{"Bit rate":"1340"},"uses_additional_data":true},{"leaderboard":"/sota/image-compression-on-cars-196","task":"Image Compression","dataset":"Cars-196","model":"Lossyless Compressor","rank_in_archive_order":1,"of":1,"metrics":{"Bit rate":"1470"},"uses_additional_data":true},{"leaderboard":"/sota/image-compression-on-food-101","task":"Image Compression","dataset":"Food-101","model":"Lossyless Compressor","rank_in_archive_order":1,"of":1,"metrics":{"Bit rate":"1270"},"uses_additional_data":true},{"leaderboard":"/sota/image-compression-on-oxford-iiit-pets","task":"Image Compression","dataset":"Oxford-IIIT Pet Dataset","model":"Lossyless Compressor","rank_in_archive_order":1,"of":1,"metrics":{"Bit rate":"1210"},"uses_additional_data":true},{"leaderboard":"/sota/image-compression-on-pcam","task":"Image Compression","dataset":"PCam","model":"Lossyless Compressor","rank_in_archive_order":1,"of":1,"metrics":{"Bit rate":"1490"},"uses_additional_data":true},{"leaderboard":"/sota/image-compression-on-stl-10","task":"Image Compression","dataset":"STL-10","model":"Lossyless Compressor","rank_in_archive_order":1,"of":1,"metrics":{"Bit rate":"1340"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2106.10800","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2106.10800"}},"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":"deterministic:regex_extraction","url":"https://github.com/YannDubs/lossyless","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"deterministic:regex_extraction","url":"https://github.com/openai/CLIP","reach":{"status":"unanswered"}}],"summary":{"ran":1,"ran_honours":1,"unverified":9},"by_repo_kind":{"official":{"samples":11,"ran":2,"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":"bd13dca3e1023087","entry":"read_bytes","repo":"YannDubs/lossyless","repo_kind":"official","path":"hub/compressor.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/hub/compressor.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"bd13dca3e1023087"}},{"code_sha256_prefix":"ae94682e8b678a9e","entry":"read_uints","repo":"YannDubs/lossyless","repo_kind":"official","path":"hub/compressor.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/hub/compressor.py","link_basis":"plan_row","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ae94682e8b678a9e"}},{"code_sha256_prefix":"ca462457e364bdc8","entry":"batch_flatten","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/helpers.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/helpers.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ca462457e364bdc8"}},{"code_sha256_prefix":"8a9a3c914c967c7d","entry":"dict_mean","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/helpers.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/helpers.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8a9a3c914c967c7d"}},{"code_sha256_prefix":"54f70d49860ba87c","entry":"get_Normalization","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/architectures.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/architectures.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"54f70d49860ba87c"}},{"code_sha256_prefix":"e984324a4075fe28","entry":"get_distortion_estimator","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/distortions.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/distortions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e984324a4075fe28"}},{"code_sha256_prefix":"d5392321fab0a866","entry":"get_featurizer_predictor","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/predictors.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/predictors.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d5392321fab0a866"}},{"code_sha256_prefix":"ed851002ec8aabc9","entry":"get_marginalDist","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/distributions.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/distributions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ed851002ec8aabc9"}},{"code_sha256_prefix":"052ae753efea2e31","entry":"get_scale_table","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/rates.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/rates.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"052ae753efea2e31"}},{"code_sha256_prefix":"a0b0ecede0ebcf72","entry":"is_plot","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/callbacks.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/callbacks.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a0b0ecede0ebcf72"}},{"code_sha256_prefix":"297b95954cb4a65e","entry":"to_numpy","repo":"YannDubs/lossyless","repo_kind":"official","path":"lossyless/helpers.py","file_url":"https://github.com/YannDubs/lossyless/blob/HEAD/lossyless/helpers.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"297b95954cb4a65e"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}