{"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/generalization-in-diffusion-models-arises","title":"Generalization in diffusion models arises from geometry-adaptive harmonic representations","arxiv_id":"2310.02557","date":"2023-10-04","proceeding":null,"authors":["Zahra Kadkhodaie","Florentin Guth","Eero P. Simoncelli","Stéphane Mallat"],"abstract":"Deep neural networks (DNNs) trained for image denoising are able to generate high-quality samples with score-based reverse diffusion algorithms. These impressive capabilities seem to imply an escape from the curse of dimensionality, but recent reports of memorization of the training set raise the question of whether these networks are learning the \"true\" continuous density of the data. Here, we show that two DNNs trained on non-overlapping subsets of a dataset learn nearly the same score function, and thus the same density, when the number of training images is large enough. In this regime of strong generalization, diffusion-generated images are distinct from the training set, and are of high visual quality, suggesting that the inductive biases of the DNNs are well-aligned with the data density. We analyze the learned denoising functions and show that the inductive biases give rise to a shrinkage operation in a basis adapted to the underlying image. Examination of these bases reveals oscillating harmonic structures along contours and in homogeneous regions. We demonstrate that trained denoisers are inductively biased towards these geometry-adaptive harmonic bases since they arise not only when the network is trained on photographic images, but also when it is trained on image classes supported on low-dimensional manifolds for which the harmonic basis is suboptimal. Finally, we show that when trained on regular image classes for which the optimal basis is known to be geometry-adaptive and harmonic, the denoising performance of the networks is near-optimal.","url_abs":"https://arxiv.org/abs/2310.02557v3","url_pdf":"https://arxiv.org/pdf/2310.02557v3.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":"generalization-in-diffusion-models-arises","repo_url":"https://github.com/labforcomputationalvision/memorization_generalization_in_diffusion_models","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"image-denoising","task_name":"Image Denoising"},{"task_slug":"memorization","task_name":"Memorization"}],"methods":[{"method_slug":"diffusion","method_name":"Diffusion"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2310.02557","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.02557"}},"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/LabForComputationalVision/memorization_generalization_in_diffusion_models","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/labforcomputationalvision/memorization_generalization_in_diffusion_models","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":11,"unverified":5},"by_repo_kind":{"official":{"samples":16,"ran":11,"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":"eb15c404ffac080f","entry":"MSE","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/quality_metrics_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/quality_metrics_func.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"eb15c404ffac080f"}},{"code_sha256_prefix":"1aa782be31b7fc00","entry":"batch_psnr_numpy","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/quality_metrics_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/quality_metrics_func.py","link_basis":"first_harvest_node","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":"1aa782be31b7fc00"}},{"code_sha256_prefix":"d0324c885182c509","entry":"calc_jacobian","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/linear_approx.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/linear_approx.py","link_basis":"first_harvest_node","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":"d0324c885182c509"}},{"code_sha256_prefix":"98dacb3ad56237ba","entry":"calc_jacobian_rows","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/linear_approx.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/linear_approx.py","link_basis":"first_harvest_node","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":"98dacb3ad56237ba"}},{"code_sha256_prefix":"6ad8ccbe18be47c9","entry":"load_mnist","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/dataloader_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/dataloader_func.py","link_basis":"first_harvest_node","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":"6ad8ccbe18be47c9"}},{"code_sha256_prefix":"31ac25abb6e95520","entry":"load_set14","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/dataloader_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/dataloader_func.py","link_basis":"first_harvest_node","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":"31ac25abb6e95520"}},{"code_sha256_prefix":"6eb2528b9343169a","entry":"make_fft_filter_1d","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/synthetic_data_generators.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/synthetic_data_generators.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6eb2528b9343169a"}},{"code_sha256_prefix":"faa53db8ba9db059","entry":"make_fft_filter_2d","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/synthetic_data_generators.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/synthetic_data_generators.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"faa53db8ba9db059"}},{"code_sha256_prefix":"56aaa6b239a0ae60","entry":"make_fft_filter_2d_seprable","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/synthetic_data_generators.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/synthetic_data_generators.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"56aaa6b239a0ae60"}},{"code_sha256_prefix":"7ff802a286a451e5","entry":"sig_to_psnr","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/quality_metrics_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/quality_metrics_func.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7ff802a286a451e5"}},{"code_sha256_prefix":"bf7b787bf71b1055","entry":"single_image_loader","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/dataloader_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/dataloader_func.py","link_basis":"first_harvest_node","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":"bf7b787bf71b1055"}},{"code_sha256_prefix":"53ef977c62a678b6","entry":"calc_jacobian_row_MS","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/linear_approx.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/linear_approx.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":"53ef977c62a678b6"}},{"code_sha256_prefix":"561901e6303b8afc","entry":"init_BF_CNN_RF","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/model_loader_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/model_loader_func.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":"561901e6303b8afc"}},{"code_sha256_prefix":"0a2d1378af0b3c3c","entry":"init_UNet","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/model_loader_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/model_loader_func.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":"0a2d1378af0b3c3c"}},{"code_sha256_prefix":"706dea6429c294c3","entry":"initialize_network","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/model_loader_func.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/model_loader_func.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":"706dea6429c294c3"}},{"code_sha256_prefix":"4835e614b7b50339","entry":"univ_inv_sol","repo":"LabForComputationalVision/memorization_generalization_in_diffusion_models","repo_kind":"official","path":"code/algorithm_inv_prob.py","file_url":"https://github.com/LabForComputationalVision/memorization_generalization_in_diffusion_models/blob/HEAD/code/algorithm_inv_prob.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":"4835e614b7b50339"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}