{"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/when-and-how-does-clip-enable-domain-and","title":"When and How Does CLIP Enable Domain and Compositional Generalization?","arxiv_id":"2502.09507","date":"2025-02-13","proceeding":null,"authors":["Elias Kempf","Simon Schrodi","Max Argus","Thomas Brox"],"abstract":"The remarkable generalization performance of contrastive vision-language models like CLIP is often attributed to the diversity of their training distributions. However, key questions remain unanswered: Can CLIP generalize to an entirely unseen domain when trained on a diverse mixture of domains (domain generalization)? Can it generalize to unseen classes within partially seen domains (compositional generalization)? What factors affect such generalization? To answer these questions, we trained CLIP models on systematically constructed training distributions with controlled domain diversity and object class exposure. Our experiments show that domain diversity is essential for both domain and compositional generalization, yet compositional generalization can be surprisingly weaker than domain generalization when the training distribution contains a suboptimal subset of the test domain. Through data-centric and mechanistic analyses, we find that successful generalization requires learning of shared representations already in intermediate layers and shared circuitry.","url_abs":"https://arxiv.org/abs/2502.09507v1","url_pdf":"https://arxiv.org/pdf/2502.09507v1.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":[],"tasks":[{"task_slug":"diversity","task_name":"Diversity"},{"task_slug":"domain-generalization","task_name":"Domain Generalization"}],"methods":[{"method_slug":"clip","method_name":"CLIP"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2502.09507","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2502.09507"}},"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/salesforce/LAVIS","reach":{"status":"ok","spdx":"BSD-3-Clause"}},{"provenance":"deterministic:regex_extraction","url":"https://github.com/neuroexplicit-saar/discover-then-name","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"deterministic:regex_extraction","url":"https://github.com/saprmarks/feature-circuits","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"deterministic:regex_extraction","url":"https://github.com/ndif-team/nnsight","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":4,"ran_fixture":1,"ran":3,"unverified":24},"by_repo_kind":{"found_in_text":{"samples":32,"ran":8,"repositories":4}},"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":"98f385d847636a3e","entry":"basic_clean","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"clip/simple_tokenizer.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/clip/simple_tokenizer.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"98f385d847636a3e"}},{"code_sha256_prefix":"7d214655954e2bc6","entry":"getAttMap","repo":"salesforce/LAVIS","repo_kind":"found_in_text","path":"lavis/common/gradcam.py","file_url":"https://github.com/salesforce/LAVIS/blob/HEAD/lavis/common/gradcam.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"7d214655954e2bc6"}},{"code_sha256_prefix":"d919ae32e5e4e616","entry":"get_pairs","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"clip/simple_tokenizer.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/clip/simple_tokenizer.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":"d919ae32e5e4e616"}},{"code_sha256_prefix":"c6ec173f19f5c34d","entry":"interpolate_pos_embed","repo":"salesforce/LAVIS","repo_kind":"found_in_text","path":"lavis/models/vit.py","file_url":"https://github.com/salesforce/LAVIS/blob/HEAD/lavis/models/vit.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"c6ec173f19f5c34d"}},{"code_sha256_prefix":"98589643273920ba","entry":"main_process","repo":"salesforce/LAVIS","repo_kind":"found_in_text","path":"lavis/common/dist_utils.py","file_url":"https://github.com/salesforce/LAVIS/blob/HEAD/lavis/common/dist_utils.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"98589643273920ba"}},{"code_sha256_prefix":"791c72070c0b1cfb","entry":"node_to_dict","repo":"salesforce/LAVIS","repo_kind":"found_in_text","path":"lavis/common/config.py","file_url":"https://github.com/salesforce/LAVIS/blob/HEAD/lavis/common/config.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"791c72070c0b1cfb"}},{"code_sha256_prefix":"cb33571427334815","entry":"tile","repo":"salesforce/LAVIS","repo_kind":"found_in_text","path":"lavis/models/base_model.py","file_url":"https://github.com/salesforce/LAVIS/blob/HEAD/lavis/models/base_model.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"cb33571427334815"}},{"code_sha256_prefix":"9542161e9640b858","entry":"whitespace_clean","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"clip/simple_tokenizer.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/clip/simple_tokenizer.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9542161e9640b858"}},{"code_sha256_prefix":"0ec9fc2025c16f65","entry":"all_gather_with_grad","repo":"salesforce/LAVIS","repo_kind":"found_in_text","path":"lavis/models/base_model.py","file_url":"https://github.com/salesforce/LAVIS/blob/HEAD/lavis/models/base_model.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"0ec9fc2025c16f65"}},{"code_sha256_prefix":"013ac77858f729a5","entry":"apply","repo":"ndif-team/nnsight","repo_kind":"found_in_text","path":"src/nnsight/util.py","file_url":"https://github.com/ndif-team/nnsight/blob/HEAD/src/nnsight/util.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":"013ac77858f729a5"}},{"code_sha256_prefix":"3662e54c46137680","entry":"build_model","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"clip/model.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/clip/model.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":"3662e54c46137680"}},{"code_sha256_prefix":"6db205fad012bb49","entry":"capture_exception","repo":"ndif-team/nnsight","repo_kind":"found_in_text","path":"src/nnsight/intervention/errors.py","file_url":"https://github.com/ndif-team/nnsight/blob/HEAD/src/nnsight/intervention/errors.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":"6db205fad012bb49"}},{"code_sha256_prefix":"4426e27aa118f08d","entry":"concat","repo":"ndif-team/nnsight","repo_kind":"found_in_text","path":"src/nnsight/intervention/batching.py","file_url":"https://github.com/ndif-team/nnsight/blob/HEAD/src/nnsight/intervention/batching.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":"4426e27aa118f08d"}},{"code_sha256_prefix":"643539116e049080","entry":"download_cached_file","repo":"salesforce/LAVIS","repo_kind":"found_in_text","path":"lavis/common/dist_utils.py","file_url":"https://github.com/salesforce/LAVIS/blob/HEAD/lavis/common/dist_utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"643539116e049080"}},{"code_sha256_prefix":"5a59e925d9bdd730","entry":"from_import_path","repo":"ndif-team/nnsight","repo_kind":"found_in_text","path":"src/nnsight/util.py","file_url":"https://github.com/ndif-team/nnsight/blob/HEAD/src/nnsight/util.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":"5a59e925d9bdd730"}},{"code_sha256_prefix":"07b539dc323f7bf6","entry":"get_annotation","repo":"saprmarks/feature-circuits","repo_kind":"found_in_text","path":"data_loading_utils.py","file_url":"https://github.com/saprmarks/feature-circuits/blob/HEAD/data_loading_utils.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":"07b539dc323f7bf6"}},{"code_sha256_prefix":"78a00cb43e4b428b","entry":"get_name_gemma","repo":"saprmarks/feature-circuits","repo_kind":"found_in_text","path":"circuit_plotting.py","file_url":"https://github.com/saprmarks/feature-circuits/blob/HEAD/circuit_plotting.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":"78a00cb43e4b428b"}},{"code_sha256_prefix":"77dfb02ffefe5bbc","entry":"get_name_pythia","repo":"saprmarks/feature-circuits","repo_kind":"found_in_text","path":"circuit_plotting.py","file_url":"https://github.com/saprmarks/feature-circuits/blob/HEAD/circuit_plotting.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":"77dfb02ffefe5bbc"}},{"code_sha256_prefix":"0b14ce4fe263ff5b","entry":"get_probe_classifier_ckpt","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"dncbm/utils.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/dncbm/utils.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":"0b14ce4fe263ff5b"}},{"code_sha256_prefix":"6d0212913db960d6","entry":"get_sae_ckpt","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"dncbm/utils.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/dncbm/utils.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":"6d0212913db960d6"}},{"code_sha256_prefix":"a31e3e8047f3f679","entry":"is_model_running","repo":"ndif-team/nnsight","repo_kind":"found_in_text","path":"src/nnsight/ndif.py","file_url":"https://github.com/ndif-team/nnsight/blob/HEAD/src/nnsight/ndif.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":"a31e3e8047f3f679"}},{"code_sha256_prefix":"b91b981015bbf498","entry":"load","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"clip/clip.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/clip/clip.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":"b91b981015bbf498"}},{"code_sha256_prefix":"900950517fc5e4d5","entry":"load_examples_nopair","repo":"saprmarks/feature-circuits","repo_kind":"found_in_text","path":"data_loading_utils.py","file_url":"https://github.com/saprmarks/feature-circuits/blob/HEAD/data_loading_utils.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":"900950517fc5e4d5"}},{"code_sha256_prefix":"6a750256744289a9","entry":"ndif_status","repo":"ndif-team/nnsight","repo_kind":"found_in_text","path":"src/nnsight/ndif.py","file_url":"https://github.com/ndif-team/nnsight/blob/HEAD/src/nnsight/ndif.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":"6a750256744289a9"}},{"code_sha256_prefix":"a7b85fe417248ff9","entry":"plot_circuit_posaligned","repo":"saprmarks/feature-circuits","repo_kind":"found_in_text","path":"circuit_plotting.py","file_url":"https://github.com/saprmarks/feature-circuits/blob/HEAD/circuit_plotting.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":"a7b85fe417248ff9"}},{"code_sha256_prefix":"fd2e79ac3e8d7ce8","entry":"sparse_flatten","repo":"saprmarks/feature-circuits","repo_kind":"found_in_text","path":"coo_utils.py","file_url":"https://github.com/saprmarks/feature-circuits/blob/HEAD/coo_utils.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":"fd2e79ac3e8d7ce8"}},{"code_sha256_prefix":"5d898b6d7272c096","entry":"sparse_mean","repo":"saprmarks/feature-circuits","repo_kind":"found_in_text","path":"coo_utils.py","file_url":"https://github.com/saprmarks/feature-circuits/blob/HEAD/coo_utils.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":"5d898b6d7272c096"}},{"code_sha256_prefix":"eaebec17f584ac1f","entry":"sparse_reshape","repo":"saprmarks/feature-circuits","repo_kind":"found_in_text","path":"coo_utils.py","file_url":"https://github.com/saprmarks/feature-circuits/blob/HEAD/coo_utils.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":"eaebec17f584ac1f"}},{"code_sha256_prefix":"fb66304fd81f1b4a","entry":"status","repo":"ndif-team/nnsight","repo_kind":"found_in_text","path":"src/nnsight/ndif.py","file_url":"https://github.com/ndif-team/nnsight/blob/HEAD/src/nnsight/ndif.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":"fb66304fd81f1b4a"}},{"code_sha256_prefix":"c3262cc8bbc0cf2f","entry":"to_import_path","repo":"ndif-team/nnsight","repo_kind":"found_in_text","path":"src/nnsight/util.py","file_url":"https://github.com/ndif-team/nnsight/blob/HEAD/src/nnsight/util.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":"c3262cc8bbc0cf2f"}},{"code_sha256_prefix":"77c07376a0e1c1a0","entry":"to_numpy","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"dncbm/visualization/drawing_functions.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/dncbm/visualization/drawing_functions.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":"77c07376a0e1c1a0"}},{"code_sha256_prefix":"abb04880ee7994e0","entry":"to_numpy_img","repo":"neuroexplicit-saar/discover-then-name","repo_kind":"found_in_text","path":"dncbm/visualization/drawing_functions.py","file_url":"https://github.com/neuroexplicit-saar/discover-then-name/blob/HEAD/dncbm/visualization/drawing_functions.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":"abb04880ee7994e0"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}