{"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/toward-a-visual-concept-vocabulary-for-gan-1","title":"Toward a Visual Concept Vocabulary for GAN Latent Space","arxiv_id":"2110.04292","date":"2021-10-08","proceeding":"ICCV 2021 10","authors":["Sarah Schwettmann","Evan Hernandez","David Bau","Samuel Klein","Jacob Andreas","Antonio Torralba"],"abstract":"A large body of recent work has identified transformations in the latent spaces of generative adversarial networks (GANs) that consistently and interpretably transform generated images. But existing techniques for identifying these transformations rely on either a fixed vocabulary of pre-specified visual concepts, or on unsupervised disentanglement techniques whose alignment with human judgments about perceptual salience is unknown. This paper introduces a new method for building open-ended vocabularies of primitive visual concepts represented in a GAN's latent space. Our approach is built from three components: (1) automatic identification of perceptually salient directions based on their layer selectivity; (2) human annotation of these directions with free-form, compositional natural language descriptions; and (3) decomposition of these annotations into a visual concept vocabulary, consisting of distilled directions labeled with single words. Experiments show that concepts learned with our approach are reliable and composable -- generalizing across classes, contexts, and observers, and enabling fine-grained manipulation of image style and content.","url_abs":"https://arxiv.org/abs/2110.04292v1","url_pdf":"https://arxiv.org/pdf/2110.04292v1.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":"toward-a-visual-concept-vocabulary-for-gan-1","repo_url":"https://github.com/schwettmann/visual-vocab","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"disentanglement","task_name":"Disentanglement"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2110.04292","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2110.04292"}},"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/schwettmann/visual-vocab","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_honours":1,"ran_draft_wrong":2,"unverified":16},"by_repo_kind":{"official":{"samples":19,"ran":3,"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":"e6e2771a0de59277","entry":"apply_ablation_replacement","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/nethook.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/nethook.py","link_basis":"harvester_set","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e6e2771a0de59277"}},{"code_sha256_prefix":"0a92af765d5132e3","entry":"make_matching_tensor","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/nethook.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/nethook.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0a92af765d5132e3"}},{"code_sha256_prefix":"aa2af4041a7e30bc","entry":"subsequence","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/nethook.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/nethook.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"aa2af4041a7e30bc"}},{"code_sha256_prefix":"c58beb7c0ad92d3b","entry":"border_from_mask","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/imgviz.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/imgviz.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":"c58beb7c0ad92d3b"}},{"code_sha256_prefix":"cfd108b7ac813d88","entry":"data_attrs","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/labwidget.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/labwidget.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":"cfd108b7ac813d88"}},{"code_sha256_prefix":"b45785074318f58d","entry":"default_loader","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/parallelfolder.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/parallelfolder.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":"b45785074318f58d"}},{"code_sha256_prefix":"820d9a91ff407527","entry":"expand_last_filename","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/imgsave.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/imgsave.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":"820d9a91ff407527"}},{"code_sha256_prefix":"7d3d2f06691b22be","entry":"grayscale_loader","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/parallelfolder.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/parallelfolder.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":"7d3d2f06691b22be"}},{"code_sha256_prefix":"606f24a2e8f3b10e","entry":"is_npy_file","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/parallelfolder.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/parallelfolder.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":"606f24a2e8f3b10e"}},{"code_sha256_prefix":"6130cee0b8ea8193","entry":"layer_regularizer_L2","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/generate_lsds.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/generate_lsds.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":"6130cee0b8ea8193"}},{"code_sha256_prefix":"1ef79e9ba4d9be9c","entry":"layer_regularizer_l1","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/generate_lsds.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/generate_lsds.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":"1ef79e9ba4d9be9c"}},{"code_sha256_prefix":"c4b1e3a489d47780","entry":"load_json","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/easydict.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/easydict.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":"c4b1e3a489d47780"}},{"code_sha256_prefix":"d7aa10473bd84c24","entry":"make_svg_bargraph","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/bargraph.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/bargraph.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":"d7aa10473bd84c24"}},{"code_sha256_prefix":"5304f2f28e1e74ac","entry":"minify","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/labwidget.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/labwidget.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":"5304f2f28e1e74ac"}},{"code_sha256_prefix":"bdd83c4ac66fe892","entry":"num_items","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/imgsave.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/imgsave.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":"bdd83c4ac66fe892"}},{"code_sha256_prefix":"4888af493ac14d04","entry":"pidfile_taken","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/pidfile.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/pidfile.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":"4888af493ac14d04"}},{"code_sha256_prefix":"67869aa0e58a1c8e","entry":"strip_image_from_grid_row","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/imgviz.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/imgviz.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":"67869aa0e58a1c8e"}},{"code_sha256_prefix":"45d86c3f32a219f8","entry":"style_attr","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/labwidget.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/labwidget.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":"45d86c3f32a219f8"}},{"code_sha256_prefix":"b72ea5e7a549503a","entry":"tqdm_terminal","repo":"schwettmann/visual-vocab","repo_kind":"official","path":"visualvocab/deps/netdissect/pbar.py","file_url":"https://github.com/schwettmann/visual-vocab/blob/HEAD/visualvocab/deps/netdissect/pbar.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":"b72ea5e7a549503a"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}