{"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/memorization-precedes-generation-learning","title":"Memorization Precedes Generation: Learning Unsupervised GANs with Memory Networks","arxiv_id":"1803.01500","date":"2018-03-05","proceeding":"ICLR 2018 1","authors":["Youngjin Kim","Minjung Kim","Gunhee Kim"],"abstract":"We propose an approach to address two issues that commonly occur during\ntraining of unsupervised GANs. First, since GANs use only a continuous latent\ndistribution to embed multiple classes or clusters of data, they often do not\ncorrectly handle the structural discontinuity between disparate classes in a\nlatent space. Second, discriminators of GANs easily forget about past generated\nsamples by generators, incurring instability during adversarial training. We\nargue that these two infamous problems of unsupervised GAN training can be\nlargely alleviated by a learnable memory network to which both generators and\ndiscriminators can access. Generators can effectively learn representation of\ntraining samples to understand underlying cluster distributions of data, which\nease the structure discontinuity problem. At the same time, discriminators can\nbetter memorize clusters of previously generated samples, which mitigate the\nforgetting problem. We propose a novel end-to-end GAN model named memoryGAN,\nwhich involves a memory network that is unsupervisedly trainable and integrable\nto many existing GAN models. With evaluations on multiple datasets such as\nFashion-MNIST, CelebA, CIFAR10, and Chairs, we show that our model is\nprobabilistically interpretable, and generates realistic image samples of high\nvisual fidelity. The memoryGAN also achieves the state-of-the-art inception\nscores over unsupervised GAN models on the CIFAR10 dataset, without any\noptimization tricks and weaker divergences.","url_abs":"http://arxiv.org/abs/1803.01500v2","url_pdf":"http://arxiv.org/pdf/1803.01500v2.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":"memorization-precedes-generation-learning","repo_url":"https://github.com/whyjay/memoryGAN","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"memorization","task_name":"Memorization"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"memory-network","method_name":"Memory Network"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1803.01500","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1803.01500"}},"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/whyjay/memoryGAN","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":6},"by_repo_kind":{"official":{"samples":6,"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":"41140f56fd26f5f3","entry":"base_d","repo":"whyjay/memoryGAN","repo_kind":"official","path":"models/discriminator.py","file_url":"https://github.com/whyjay/memoryGAN/blob/HEAD/models/discriminator.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":"41140f56fd26f5f3"}},{"code_sha256_prefix":"680af30724829643","entry":"base_g","repo":"whyjay/memoryGAN","repo_kind":"official","path":"models/generator.py","file_url":"https://github.com/whyjay/memoryGAN/blob/HEAD/models/generator.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":"680af30724829643"}},{"code_sha256_prefix":"cd2dd7d4d3a3dff6","entry":"dense_to_one_hot","repo":"whyjay/memoryGAN","repo_kind":"official","path":"affmnist.py","file_url":"https://github.com/whyjay/memoryGAN/blob/HEAD/affmnist.py","link_basis":"harvester_set","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":"cd2dd7d4d3a3dff6"}},{"code_sha256_prefix":"24b10f0c7a83f999","entry":"extract_images","repo":"whyjay/memoryGAN","repo_kind":"official","path":"affmnist.py","file_url":"https://github.com/whyjay/memoryGAN/blob/HEAD/affmnist.py","link_basis":"harvester_set","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":"24b10f0c7a83f999"}},{"code_sha256_prefix":"1318c92a86283291","entry":"extract_labels","repo":"whyjay/memoryGAN","repo_kind":"official","path":"affmnist.py","file_url":"https://github.com/whyjay/memoryGAN/blob/HEAD/affmnist.py","link_basis":"harvester_set","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":"1318c92a86283291"}},{"code_sha256_prefix":"82ce97ede63f1724","entry":"memory_d","repo":"whyjay/memoryGAN","repo_kind":"official","path":"models/discriminator.py","file_url":"https://github.com/whyjay/memoryGAN/blob/HEAD/models/discriminator.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":"82ce97ede63f1724"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}