{"doi":"10.52202/079017-0749","title":"Cross-Care: Assessing the Healthcare Implications of Pre-training Data on Language Model Bias","abstract":"influence the outputs of LLMs. We expose and quantify discrepancies by juxtaposing these biases against actual disease prevalences in various U.S. demographic groups. Our results highlight substantial misalignment between LLM representation of disease prevalence and real disease prevalence rates across demographic subgroups, indicating a pronounced risk of bias propagation and a lack of real-world grounding for medical applications of LLMs. Furthermore, we observe that various alignment methods minimally resolve inconsistencies in the models' representation of disease prevalence across different languages. For further exploration and analysis, we make all data and a data visualization tool available at: www.crosscare.net.","journal":"PubMed","year":2024,"id":484548,"datarank":0.0,"base_score":0.0,"endowment":0.0,"self_citation_contribution":0.0,"citation_network_contribution":0.0,"self_endowment_contribution":0.0,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":9,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":0.9455,"is_data_producer":false,"deposit_databanks":null,"is_oa":true,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":"2024-01-01","fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":648059,"name":"Jack Gallifant","orcid":"0000-0003-1306-2334","position":1,"is_corresponding":false},{"id":1326320,"name":"Mingye Gao","orcid":null,"position":2,"is_corresponding":false},{"id":1325737,"name":"Pedro Moreira","orcid":"0000-0002-7035-7799","position":3,"is_corresponding":false},{"id":1326319,"name":"Nikolaj Munch","orcid":null,"position":4,"is_corresponding":false},{"id":1327190,"name":"Ajay Muthukkumar","orcid":null,"position":5,"is_corresponding":false},{"id":1326721,"name":"Arvind Rajan","orcid":"0000-0003-4829-5007","position":6,"is_corresponding":false},{"id":1327191,"name":"Jaya Kolluri","orcid":null,"position":7,"is_corresponding":false},{"id":1221633,"name":"Amelia Fiske","orcid":"0000-0001-7207-6897","position":8,"is_corresponding":false},{"id":71,"name":"Janna Hastings","orcid":"0000-0002-3469-4923","position":9,"is_corresponding":false},{"id":37161,"name":"Hugo J.W.L. Aerts","orcid":"0000-0002-2122-2003","position":10,"is_corresponding":false},{"id":422836,"name":"Brian Anthony","orcid":"0000-0001-6346-5276","position":11,"is_corresponding":false},{"id":4662,"name":"Leo Anthony Celi","orcid":"0000-0001-6712-6626","position":12,"is_corresponding":false},{"id":469461,"name":"William La Cava","orcid":"0000-0002-1332-2960","position":13,"is_corresponding":false},{"id":229099,"name":"Danielle S. Bitterman","orcid":"0000-0003-0345-2232","position":14,"is_corresponding":false},{"id":1164181,"name":"Shan Chen","orcid":"0000-0003-0912-7254","position":0,"is_corresponding":true}],"reference_count":0,"raw_metadata":null,"created_at":"2026-07-19T02:07:42.971417Z","pmid":"40313683","pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}