{"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/facts-first-amplify-correlations-and-then-1","title":"FACTS: First Amplify Correlations and Then Slice to Discover Bias","arxiv_id":"2309.17430","date":"2023-09-29","proceeding":"ICCV 2023 1","authors":["Sriram Yenamandra","Pratik Ramesh","Viraj Prabhu","Judy Hoffman"],"abstract":"Computer vision datasets frequently contain spurious correlations between task-relevant labels and (easy to learn) latent task-irrelevant attributes (e.g. context). Models trained on such datasets learn \"shortcuts\" and underperform on bias-conflicting slices of data where the correlation does not hold. In this work, we study the problem of identifying such slices to inform downstream bias mitigation strategies. We propose First Amplify Correlations and Then Slice to Discover Bias (FACTS), wherein we first amplify correlations to fit a simple bias-aligned hypothesis via strongly regularized empirical risk minimization. Next, we perform correlation-aware slicing via mixture modeling in bias-aligned feature space to discover underperforming data slices that capture distinct correlations. Despite its simplicity, our method considerably improves over prior work (by as much as 35% precision@10) in correlation bias identification across a range of diverse evaluation settings. Our code is available at: https://github.com/yvsriram/FACTS.","url_abs":"https://arxiv.org/abs/2309.17430v1","url_pdf":"https://arxiv.org/pdf/2309.17430v1.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":"facts-first-amplify-correlations-and-then-1","repo_url":"https://github.com/yvsriram/facts","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2309.17430","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2309.17430"}},"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/yvsriram/facts","reach":{"status":"ok"}},{"provenance":"deterministic:regex_extraction","url":"https://github.com/yvsriram/FACTS","reach":{"status":"ok"}}],"summary":{"ran":10,"ran_draft_wrong":1,"unverified":3},"by_repo_kind":{"official":{"samples":14,"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":14,"samples":[{"code_sha256_prefix":"123034cc3c4b573c","entry":"compute_class_wise_retrieval","repo":"yvsriram/FACTS","repo_kind":"official","path":"select_best_point.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/select_best_point.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"123034cc3c4b573c"}},{"code_sha256_prefix":"fc0baacecd6328e9","entry":"compute_overall_retrieval","repo":"yvsriram/FACTS","repo_kind":"official","path":"select_best_point.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/select_best_point.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"fc0baacecd6328e9"}},{"code_sha256_prefix":"e33b112514a9e4b0","entry":"generate2","repo":"yvsriram/FACTS","repo_kind":"official","path":"CoSi/_describe/generate_captions.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/CoSi/_describe/generate_captions.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"e33b112514a9e4b0"}},{"code_sha256_prefix":"883320eb62f15d12","entry":"generate_beam","repo":"yvsriram/FACTS","repo_kind":"official","path":"CoSi/_describe/generate_captions.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/CoSi/_describe/generate_captions.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"883320eb62f15d12"}},{"code_sha256_prefix":"e1fe6b1f19460a65","entry":"get_device","repo":"yvsriram/FACTS","repo_kind":"official","path":"CoSi/_describe/generate_captions.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/CoSi/_describe/generate_captions.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"e1fe6b1f19460a65"}},{"code_sha256_prefix":"f9d167bde46e08be","entry":"get_loaders","repo":"yvsriram/FACTS","repo_kind":"official","path":"utils/datasets.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/utils/datasets.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f9d167bde46e08be"}},{"code_sha256_prefix":"c0769dd0245751eb","entry":"get_pairwise_similarity","repo":"yvsriram/FACTS","repo_kind":"official","path":"utils/misc.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/utils/misc.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"c0769dd0245751eb"}},{"code_sha256_prefix":"545b08bf27adbd01","entry":"get_wrapped_datasets","repo":"yvsriram/FACTS","repo_kind":"official","path":"utils/datasets.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/utils/datasets.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"545b08bf27adbd01"}},{"code_sha256_prefix":"deaf75ee14d1af3b","entry":"label_to_class_mapping","repo":"yvsriram/FACTS","repo_kind":"official","path":"slice.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/slice.py","link_basis":"plan_row","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"deaf75ee14d1af3b"}},{"code_sha256_prefix":"e91f020101b7a8a6","entry":"requires_columns","repo":"yvsriram/FACTS","repo_kind":"official","path":"CoSi/utils.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/CoSi/utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"e91f020101b7a8a6"}},{"code_sha256_prefix":"f825049be1c09b4c","entry":"row_norms","repo":"yvsriram/FACTS","repo_kind":"official","path":"CoSi/_slice/mixture.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/CoSi/_slice/mixture.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f825049be1c09b4c"}},{"code_sha256_prefix":"93ccf5ee9f74c640","entry":"get_labels_and_contexts","repo":"yvsriram/FACTS","repo_kind":"official","path":"select_best_point.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/select_best_point.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"93ccf5ee9f74c640"}},{"code_sha256_prefix":"4ec6f4a3c612770f","entry":"nested_getattr","repo":"yvsriram/FACTS","repo_kind":"official","path":"CoSi/utils.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/CoSi/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"4ec6f4a3c612770f"}},{"code_sha256_prefix":"9ee49e11f4ea1f78","entry":"plot_panel","repo":"yvsriram/FACTS","repo_kind":"official","path":"CoSi/plotting_utils.py","file_url":"https://github.com/yvsriram/FACTS/blob/HEAD/CoSi/plotting_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9ee49e11f4ea1f78"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}