{"as_of":"2026-08-22T04:51:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:0cf0d41fff7246757b44ed1a8a9b30778d0f3348ce88def8486f7faeba46fbe2","coverage":[{"denominator":49,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":49,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-02T02:54:04.303701Z","state":"measured"},{"denominator":49,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":49,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-21T06:32:19.484+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2607.14205/citation-record","integrity":"/paper/2607.14205/integrity","json":"/paper/2607.14205/citation-record.json","paper":"/paper/2607.14205"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:53:59.601627Z","title":"The effect of clinical information on radiology reporting: A systematic review","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-02T02:53:59.601627Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:2ee6934ad44eda2305330b8d69e58fbfeb320b99bc6bfa95d54dfdfffc5fcfcd","observation_id":"3a15211c-2686-45de-ae3e-db33aa989642","resolution":{"observed_at":"2026-08-02T02:53:59.601627Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:53:59.660071Z","title":"Large language models in medicine","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-02T02:53:59.660071Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:f705b64553bc41850dddeb811cfb6ea4e00339e974e6a08879a2bcc647b186ee","observation_id":"2b5b491c-6984-4a5f-8d33-17aab07b3d03","resolution":{"observed_at":"2026-08-02T02:53:59.660071Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:53:59.772144Z","title":"Improving language understanding by generative pre-training","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-02T02:53:59.772144Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:f0f1c62e31ccd6e626bef215cab9125db44a0a0013e9480133d4351c436d1f16","observation_id":"770c787c-d3dc-4733-9b18-a18c880d328b","resolution":{"observed_at":"2026-08-02T02:53:59.772144Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:53:59.868498Z","title":"Revolutionizing radiology with GPT-based models: current ap- plications, future possibilities and limitations of ChatGPT","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-02T02:53:59.868498Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:bfb1787c6ba13e74f02d31ebf87c11cbe69c6865525c471d2511e846e2154a05","observation_id":"6166fded-c395-453b-af47-6f2997266cff","resolution":{"observed_at":"2026-08-02T02:53:59.868498Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":"10.2196/65547","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Large language models in summarizing radiology re- port impressions for lung cancer in Chinese: evaluation study","venue":"Journal of Medical Internet Research","work_id":"3d4986c1-d8ae-4388-a6bc-c9e47b8bca36","year":2025},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-02T02:53:59.997026Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:527179965aba5d0a8bb9f8caafcaffddfb8632cc920c45d05e71095641856bc1","observation_id":"3f1f4ce7-227c-43df-be29-b5fbd0669d23","resolution":{"observed_at":"2026-08-02T02:59:31.714580Z","resolver_source":"doi_truncated","status":"malformed_identifier"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":"10.2196/55799","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Evaluating large language models for automated reporting and data systems categorization: cross-sectional study","venue":"JMIR Medical Informatics","work_id":"be23cdc0-38e6-47e5-be38-6e4f109b702a","year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.131231Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:7da726174448120ac425dad315208ed2181ba2bd534e84d13ee8cc40cf1ddf37","observation_id":"b519bfa8-404f-4482-8288-681011312bf6","resolution":{"observed_at":"2026-08-02T02:59:31.577310Z","resolver_source":"doi_truncated","status":"malformed_identifier"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:00.261937Z","title":"A systematic review of large language model (LLM) evaluations in clinical medicine","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.261937Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:602274e175065ffac9e423acebe8913e3f355ff799da00fb3113830bf621fca8","observation_id":"68bbfe0a-fec3-41d1-baa5-f9780532e3ae","resolution":{"observed_at":"2026-08-02T02:54:00.261937Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:00.380164Z","title":"The HIPAA Privacy Rule","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.380164Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:e3cbf948c46368922b7f93392a8bc17d9a82063c015b84cad05975650bef7674","observation_id":"4254fa76-84d4-4530-bb2c-a8b9f66b62aa","resolution":{"observed_at":"2026-08-02T02:54:00.380164Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:00.468323Z","title":"General Data Protection Regulation (GDPR)","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.468323Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:244adfcdebcc18b2b736026e1cbae47596b5579d0eedee3a55a780731520d934","observation_id":"b75662c2-0f13-4ec2-8c8d-16d2ed40c48a","resolution":{"observed_at":"2026-08-02T02:54:00.468323Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:00.564045Z","title":"Communication-efficient learning of deep networks from decentralized data","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.564045Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:e8622143bee9cfa4e97914147fb13b090a69e8cff718669d77b13ff8eb5eb941","observation_id":"41424005-23fa-438b-ab77-d391ccf01add","resolution":{"observed_at":"2026-08-02T02:54:00.564045Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:00.660131Z","title":"Continually tuning a large language model for multi-domain radiology report generation","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.660131Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:31c286fb1c6d88854ee82abddb413aff8e101b1f9935ababe586d89c1cb31fbc","observation_id":"95721040-a1a3-42a5-aaca-f7e1190367ff","resolution":{"observed_at":"2026-08-02T02:54:00.660131Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:00.784989Z","title":"Chatbots and large language models in radiology: a practical primer for clinical and research applications","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.784989Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:d2c4d460f96bc20328e50aa516715f65b677b6055b7be3dc23ba306cc4f7820c","observation_id":"35f503c1-de95-4dda-b66a-99f0cfafc665","resolution":{"observed_at":"2026-08-02T02:54:00.784989Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:00.891232Z","title":"Large language models in radiology: fundamentals, applications, ethical considerations, risks, and future directions","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.891232Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:f2b3fec0903e9eb0b2876a9dedddd649ddbbad6589fed2af6d30a2af89feb90b","observation_id":"34d5e64f-ec31-42f5-b98a-b963665cd200","resolution":{"observed_at":"2026-08-02T02:54:00.891232Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:00.979914Z","title":"Leveraging foundation and large language models in medical artificial intelligence","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:00.979914Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:b7fadf44c85f43ebf0ba4c00906bbbd9404926743c9c5ca880d7af688adcbfbb","observation_id":"4aa56851-686f-4424-b7d8-5a0283bc0cd8","resolution":{"observed_at":"2026-08-02T02:54:00.979914Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:01.078758Z","title":"Client security alone fails in federated learning: 2D and 3D attack insights","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.078758Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:5cbeec482c43d701fcfe6825ccf5042d0213df3ee3046d083b6dd32df473376d","observation_id":"be52aa86-da7a-451e-a47d-69ca3b2e7db5","resolution":{"observed_at":"2026-08-02T02:54:01.078758Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:01.146651Z","title":"Federated machine learning, privacy-enhancing technologies, and 26 data protection laws in medical research: scoping review","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.146651Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:9c32e4a883025faae51bc6d50680d4beea00ec728e1d61eecc8eaa395a594c0b","observation_id":"d86de250-1d5d-4ddd-bcfd-80a6a2fa5b2d","resolution":{"observed_at":"2026-08-02T02:54:01.146651Z","resolver_source":null,"status":"malformed_identifier"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:01.211205Z","title":"Personalized and privacy- preserving federated heterogeneous medical image analysis with PPPML-HMI","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.211205Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:677bf827f5fdfa483f5b40c91b3eec9de133d8c41c7bbb5973138a96eb6b1772","observation_id":"c4f6e849-2c52-424d-8361-805032131afe","resolution":{"observed_at":"2026-08-02T02:54:01.211205Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:01.282516Z","title":"End-to-end privacy preserving deep learning on multi-institutional medical imaging","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.282516Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:f08fe31d38840355b1ae4b837143f32596932d32cf5eb60b0355acfda391af03","observation_id":"b145b988-5daa-4553-9021-98e3dc95f195","resolution":{"observed_at":"2026-08-02T02:54:01.282516Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:01.349480Z","title":"Do gradient inversion attacks make federated learning unsafe? IEEE Trans Med Imaging IEEE; 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.349480Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:2796a1a58a90c7ca3e884f42256d4b7f7090b21986e4da28c5481a470ed798b8","observation_id":"ff834ef5-a860-47dd-abb7-cff17070025d","resolution":{"observed_at":"2026-08-02T02:54:01.349480Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:01.444375Z","title":"A new era of text mining in radiology with privacy- preserving LLMs","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.444375Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:8e6332aad6a8a0cb35e5f63e26c7a0dbeffc3dd84157d48439053590058c6cff","observation_id":"18f20ab1-983c-4342-845b-aa67020d42f8","resolution":{"observed_at":"2026-08-02T02:54:01.444375Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:01.536181Z","title":"Domain- specific language model pretraining for biomedical natural language processing","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.536181Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:2a1fa37276087683ee66f458efb3d78217d73974d77c04d3f709b3e2d1d32a26","observation_id":"df8d18c0-da87-4f85-9f91-301b6f926719","resolution":{"observed_at":"2026-08-02T02:54:01.536181Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.14558","last_updated":"2025-03-31T21:04:11Z","snapshot_observed_at":"2026-08-16T14:49:54.840889Z","submitted_at":"2023-10-23T04:22:50Z","title":"AlpaCare:Instruction-tuned Large Language Models for Medical Application","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.14558","snapshot_observed_at":"2026-08-02T02:54:01.686176Z","title":"arXiv preprint arXiv:2310.14558; 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.686176Z"},"links":{"cited_paper":"/paper/2310.14558","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:326fcaf4c52392e7a191d70b0ed9b7d415961366385e31ea80680ef44ca574b3","observation_id":"c827d9c3-7cb6-45be-83ae-c3ca344dbe61","resolution":{"observed_at":"2026-08-02T02:54:01.686176Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:01.808689Z","title":"RadiolArtifIntell2022July1;4(4):e210258","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:01.808689Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:889a97901c6e1e3ad4a0f7a736ff7335d4218500c2000271e0ccf5e4483dab03","observation_id":"e4a8e9d8-05e0-4350-9674-4521b0c1bd90","resolution":{"observed_at":"2026-08-02T02:54:01.808689Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2307.09288","last_updated":"2023-07-19T17:08:59Z","snapshot_observed_at":"2026-08-07T12:56:43.323460Z","submitted_at":"2023-07-18T14:31:57Z","title":"Llama 2: Open Foundation and Fine-Tuned Chat Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2307.09288","snapshot_observed_at":"2026-08-02T02:54:02.158327Z","title":"Llama 2: Open foundation and fine-tuned chat models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:02.158327Z"},"links":{"cited_paper":"/paper/2307.09288","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:3f63ab28f7aa1df85b3174a4f3515eac364cfb59e12dde1bf8444369ad965be6","observation_id":"6c4375ce-dcd9-44cd-99a7-22182b6178bd","resolution":{"observed_at":"2026-08-02T02:54:02.158327Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2001.02610","last_updated":"2020-01-08T16:45:09Z","snapshot_observed_at":"2026-08-20T00:01:01.707115Z","submitted_at":"2020-01-08T16:45:09Z","title":"iDLG: Improved Deep Leakage from Gradients","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2001.02610","snapshot_observed_at":"2026-08-02T02:54:02.313861Z","title":"iDLG: improved deep leakage from gradients","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:02.313861Z"},"links":{"cited_paper":"/paper/2001.02610","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:4fa9f2ac9b5a4288c23cdd728d8ddbcfb9f4dafd8f0cabbf9d2787c402fd7d9d","observation_id":"bf6ba427-6506-4db5-adca-b951ec2a362f","resolution":{"observed_at":"2026-08-02T02:54:02.313861Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2507.05201","last_updated":"2026-04-06T19:50:20Z","snapshot_observed_at":"2026-08-16T01:32:15.677523Z","submitted_at":"2025-07-07T17:01:44Z","title":"MedGemma Technical Report","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2507.05201","snapshot_observed_at":"2026-08-02T02:54:02.463765Z","title":"MedGemmatechnicalreport","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:02.463765Z"},"links":{"cited_paper":"/paper/2507.05201","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:77e838e4e24a9373dc0b818ee440a79f957cce7d1d3bcc8afffeb46153b81ba4","observation_id":"74407d5d-9115-4684-9d51-b5621d8db2f9","resolution":{"observed_at":"2026-08-02T02:54:02.463765Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:02.621947Z","title":"Dischargesum Dataset","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:02.621947Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:3594448094e04e17448eb31c2c3ed6f40f165604f33513d21b53808d007c5c76","observation_id":"681dcf08-bd6b-4651-83d9-fb1fc1bc0fbe","resolution":{"observed_at":"2026-08-02T02:54:02.621947Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1901.07042","last_updated":"2019-11-14T17:34:51Z","snapshot_observed_at":"2026-08-02T04:11:08.699777Z","submitted_at":"2019-01-21T19:01:00Z","title":"MIMIC-CXR-JPG, a large publicly available database of labeled chest radiographs","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1901.07042","snapshot_observed_at":"2026-08-02T02:54:02.741843Z","title":"MIMIC-CXR-JPG, a large publicly available database of labeled chest radiographs","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:02.741843Z"},"links":{"cited_paper":"/paper/1901.07042","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:d123f3f1327ded52c662f982dba77a74e324069fb87cd0267073afae8ea7c897","observation_id":"b982c37b-2b41-4213-8ecd-f717ed4fc376","resolution":{"observed_at":"2026-08-02T02:54:02.741843Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:02.873417Z","title":"BLEU: a method for automatic evaluation of machine translation","venue":null,"work_id":null,"year":2002},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:02.873417Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:2af2609730f694a61243cdb8c39ca87cfe38ca7644d0cdded94521a82965cea0","observation_id":"b0386864-eb48-462b-ae3f-aa10f781109d","resolution":{"observed_at":"2026-08-02T02:54:02.873417Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":"10.18653/v1/2023.bionlp-1.45","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Overview of the RadSum23 shared task on multi-modal and multi-anatomical radiology report summarization","venue":null,"work_id":"37fb76bd-5518-4926-9b15-c0daaaf0a8fe","year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:02.989708Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:69802764a2e21332e72ee5617f6daaa54e9ce7789f38389fefd9877b2f919953","observation_id":"a1f81367-e47b-47e6-9682-e5a337aba2d5","resolution":{"observed_at":"2026-08-02T02:59:31.376175Z","resolver_source":"doi","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":"10.21203/rs.3.rs-3483777/v1","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Clinical text summarization: adapting large language models can outperform human experts","venue":"Research Square","work_id":"fb7d2102-6888-4cf9-a6e2-2e515345ef64","year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.148724Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:46fe1832017c9a369f5f304bb2f3ab3e5ac382cc846b4c1b353f1251cc8dbc1e","observation_id":"9464c288-70f8-4c68-8515-3112bfa338fb","resolution":{"observed_at":"2026-08-02T02:59:31.230919Z","resolver_source":"doi","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:03.282181Z","title":"ROUGE: a package for automatic evaluation of summaries","venue":null,"work_id":null,"year":2004},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.282181Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:134c3a2b50c780120ce56745160203c591162680cf04bccffb65324e8bcb327d","observation_id":"dda8968d-2158-4d30-8def-d09eee0cd575","resolution":{"observed_at":"2026-08-02T02:54:03.282181Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:03.334390Z","title":"Differential Privacy","venue":null,"work_id":null,"year":2006},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.334390Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:92e7ca52cce303d3c0448d8cbbc88348496789d4dd89e7edfea59dda86dc3060","observation_id":"6d460acd-891d-4154-b9cc-9ca1bfc3d5a6","resolution":{"observed_at":"2026-08-02T02:54:03.334390Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:03.410429Z","title":"Practical secure aggregation for privacy-preserving machine learning","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.410429Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:b7a8f9d0873fcf3b32e516aefbb46085a4bcaa426c3049329111fc7d37282a20","observation_id":"5cf0184d-335a-4128-9b4d-d74b80d78c8f","resolution":{"observed_at":"2026-08-02T02:54:03.410429Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":"10.1145/3595185","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"ACM Trans Intell Syst Technol 2023 Dec 31;14(6):1–32","venue":"ACM Transactions on Intelligent Systems and Technology","work_id":"179ca86d-1b06-4b9c-ae09-2191b76b79b3","year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.493819Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:8e687358a7caa33c0d0f58f362d18facfd3c001af112a5699c41a9d4bc092129","observation_id":"c4e1f56d-64a7-433a-9ab4-0d3096754807","resolution":{"observed_at":"2026-08-02T02:59:31.120963Z","resolver_source":"doi","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2409.11423","last_updated":"2025-01-29T18:27:16Z","snapshot_observed_at":"2026-08-16T13:19:11.816106Z","submitted_at":"2024-09-12T10:14:12Z","title":"Generated Data with Fake Privacy: Hidden Dangers of Fine-tuning Large Language Models on Generated Data","version":2},"cited_work":{"arxiv_id":"2409.11423","doi":"10.48550/arxiv.2409.11423","metadata_source":"pith","pith_arxiv_id":"2409.11423","snapshot_observed_at":"2026-08-05T02:49:54.815029Z","title":"Generated Data with Fake Privacy: Hidden Dangers of Fine-tuning Large Language Models on Generated Data","venue":"cs.CR","work_id":"16103770-07b7-4d4f-bb36-4445424e8d22","year":2024},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.552707Z"},"links":{"cited_paper":"/paper/2409.11423","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:49dd22daa35c1228a697686d3747cab1a39ffd847d287fd6557f464e77d3cb38","observation_id":"88dbfb96-ff98-480c-a124-cde4b7442db9","resolution":{"observed_at":"2026-08-02T02:59:30.947605Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:03.604762Z","title":"Synthetic data — anonymisation groundhog day","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.604762Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:83da33ec1d299828500b4302ec6f412a1bb4ca7c1e52ebc64b8319d485ef6921","observation_id":"cce84946-9049-4d8b-8590-05360938581c","resolution":{"observed_at":"2026-08-02T02:54:03.604762Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:03.716181Z","title":"Evaluating differentially private machine learning in practice","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.716181Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:4eab1cb9bccc8515f2df88be03cfc5b630465bcbb93eed72970b8d60559a78fc","observation_id":"c5dc6ce0-ee6e-474d-af2f-5118841a6413","resolution":{"observed_at":"2026-08-02T02:54:03.716181Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2208.10919","last_updated":"2022-08-21T23:56:28Z","snapshot_observed_at":"2026-08-20T18:36:20.372304Z","submitted_at":"2022-08-21T23:56:28Z","title":"Cluster Based Secure Multi-Party Computation in Federated Learning for Histopathology Images","version":1},"cited_work":{"arxiv_id":"2208.10919","doi":"10.48550/arxiv.2208.10919","metadata_source":"pith","pith_arxiv_id":"2208.10919","snapshot_observed_at":"2026-08-05T02:49:54.815029Z","title":"Cluster Based Secure Multi-Party Computation in Federated Learning for Histopathology Images","venue":"cs.CR","work_id":"2dc43b44-bdd1-42a1-b77b-e8d1f8c0e26a","year":2022},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.868613Z"},"links":{"cited_paper":"/paper/2208.10919","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:025ff0c4d7efabed80a5984fae477a319eab77c695f3471921dce2088bafcf4d","observation_id":"f3413752-ec2e-4a46-838c-ca4542bbaa8e","resolution":{"observed_at":"2026-08-02T02:59:30.431513Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:03.921879Z","title":"A scoping review of privacy and utility metrics in medical synthetic data","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.921879Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:3a19c57666672863ec6ebcc018091794c47da6f5f74d3736ef391397b5337713","observation_id":"ed8b4867-7757-408a-b1e0-18b2a6b55e7b","resolution":{"observed_at":"2026-08-02T02:54:03.921879Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1902.08874","last_updated":"2019-08-12T23:18:20Z","snapshot_observed_at":"2026-08-22T01:09:27.101921Z","submitted_at":"2019-02-24T01:48:53Z","title":"Evaluating Differentially Private Machine Learning in Practice","version":4},"cited_work":{"arxiv_id":"1902.08874","doi":"10.48550/arxiv.1902.08874","metadata_source":"pith","pith_arxiv_id":"1902.08874","snapshot_observed_at":"2026-08-05T02:49:54.815029Z","title":"Evaluating Differentially Private Machine Learning in Practice","venue":"cs.LG","work_id":"0b5bfa00-f138-4472-b0d0-046adb3e0a47","year":2019},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.783551Z"},"links":{"cited_paper":"/paper/1902.08874","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:27f197399fb3bcee4734cc9f87f87bf3fcb24dc2ebd7b5fa89f22ad5ce02a88e","observation_id":"dd7b42aa-80a0-4b59-b280-fab069658661","resolution":{"observed_at":"2026-08-02T02:59:30.639228Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2003.14053","last_updated":"2020-09-11T11:41:10Z","snapshot_observed_at":"2026-08-15T06:18:22.062863Z","submitted_at":"2020-03-31T09:35:02Z","title":"Inverting Gradients -- How easy is it to break privacy in federated learning?","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2003.14053","snapshot_observed_at":"2026-08-02T02:54:04.056448Z","title":"Inverting gradients — how easy is it to break privacy in federated learning? Advances in Neural Information Processing Systems (NeurIPS) 2020; 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:04.056448Z"},"links":{"cited_paper":"/paper/2003.14053","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:0aba3168fe796f349e3ee4f58a860214004e032b25ae4cb892bc558d763b376b","observation_id":"363e6b18-501c-47ef-acd7-63c613010722","resolution":{"observed_at":"2026-08-02T02:54:04.056448Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:04.123412Z","title":"Advances and open problems in federated learning","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:04.123412Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:74b6454a1b2f740499dc3ae90998b87c6a1a1c9ef26c57e76eb9259d0e6d374f","observation_id":"85b70f88-2282-4d06-93ad-1d44bc669d43","resolution":{"observed_at":"2026-08-02T02:54:04.123412Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2110.13057","last_updated":"2022-03-18T05:52:49Z","snapshot_observed_at":"2026-08-18T15:26:15.761119Z","submitted_at":"2021-10-25T15:52:06Z","title":"Robbing the Fed: Directly Obtaining Private Data in Federated Learning with Modified Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2110.13057","snapshot_observed_at":"2026-08-02T02:54:03.989583Z","title":"Robbing the Fed: directly obtaining private data in federated learning with modified models","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.989583Z"},"links":{"cited_paper":"/paper/2110.13057","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:a528f57e24ea80e8a485bafb42ecd3cc0ba044ec3d337b54a5fd11cac93464ea","observation_id":"50427c0e-df0f-4ea4-90a3-fcef341eda1e","resolution":{"observed_at":"2026-08-02T02:54:03.989583Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2103.06819","last_updated":"2021-09-21T17:58:26Z","snapshot_observed_at":"2026-08-21T09:16:38.295374Z","submitted_at":"2021-03-11T17:41:32Z","title":"TAG: Gradient Attack on Transformer-based Language Models","version":6},"cited_work":{"arxiv_id":"2103.06819","doi":"10.48550/arxiv.2103.06819","metadata_source":"pith","pith_arxiv_id":"2103.06819","snapshot_observed_at":"2026-08-05T02:49:54.815029Z","title":"TAG: Gradient Attack on Transformer-based Language Models","venue":"cs.CR","work_id":"0f518453-ce7a-47c3-a138-ad5fe5232b98","year":2021},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:04.195624Z"},"links":{"cited_paper":"/paper/2103.06819","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:f152b3e1e0dca772e18870dd89a17255432fa990432d54d2cd65f55820eea5d3","observation_id":"ed58b4bc-167a-40ff-bd3e-eb99655dc2f1","resolution":{"observed_at":"2026-08-02T02:59:29.930958Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1906.08935","last_updated":"2019-12-19T05:09:29Z","snapshot_observed_at":"2026-08-14T16:12:44.382531Z","submitted_at":"2019-06-21T03:46:43Z","title":"Deep Leakage from Gradients","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1906.08935","snapshot_observed_at":"2026-08-02T02:54:04.303701Z","title":"Deep leakage from gradients","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:04.303701Z"},"links":{"cited_paper":"/paper/1906.08935","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:b0e991e3f311e046a7c452514ee71f1faa787b984f86654156d8288d518276fb","observation_id":"b62c528a-9121-47e5-b09f-91bbd9031708","resolution":{"observed_at":"2026-08-02T02:54:04.303701Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T02:54:04.181596Z","title":"When the curious abandon honesty: federated learning is not private","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:04.181596Z"},"links":{"citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:9c9c170656e3e05121a652054138e457149de09c7c4238e643807c06e54457c2","observation_id":"8ae23861-3b48-4cc3-bf3e-6b89e1d6fb8e","resolution":{"observed_at":"2026-08-02T02:54:04.181596Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1908.09203","last_updated":"2019-11-13T03:54:12Z","snapshot_observed_at":"2026-08-03T18:37:05.537362Z","submitted_at":"2019-08-24T20:41:40Z","title":"Release Strategies and the Social Impacts of Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1908.09203","snapshot_observed_at":"2026-08-02T02:54:02.028887Z","title":null,"venue":null,"work_id":null,"year":1908},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":2019,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:02.028887Z"},"links":{"cited_paper":"/paper/1908.09203","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:4d0ce3f3b3fd91d45b5707256c65a18aee762fa20f628c7fd7c4c1cfb84dca59","observation_id":"06ea29a9-bc8b-43d3-b242-097d376abe83","resolution":{"observed_at":"2026-08-02T02:54:02.028887Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2011.07018","last_updated":"2022-01-24T10:32:35Z","snapshot_observed_at":"2026-08-16T19:06:01.158800Z","submitted_at":"2020-11-13T16:58:42Z","title":"Synthetic Data -- Anonymisation Groundhog Day","version":6},"cited_work":{"arxiv_id":"2011.07018","doi":"10.48550/arxiv.2011.07018","metadata_source":"pith","pith_arxiv_id":"2011.07018","snapshot_observed_at":"2026-08-05T02:49:54.815029Z","title":"Synthetic Data -- Anonymisation Groundhog Day","venue":"cs.LG","work_id":"f82f17dd-2e79-4f40-a746-6db353ba4bc4","year":2020},"citing_paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks","version":1},"reference_index":2020,"source":"pdf_text","source_observed_at":"2026-08-02T02:54:03.657090Z"},"links":{"cited_paper":"/paper/2011.07018","citing_paper":"/paper/2607.14205"},"observation_digest":"sha256:62ff1dcb35427b7755ae6c6675c4f801dd8e8de7eb11e610793cfd81c7203e2a","observation_id":"2fe38e73-b362-4517-bf05-2eec21d6e47d","resolution":{"observed_at":"2026-08-02T02:59:30.781040Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-21T06:32:19.484+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2607.14205","last_updated":"2026-07-15T17:57:53Z","latest_version":1,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-19T07:47:28.735173Z","submitted_at":"2026-07-15T17:57:53Z","title":"Privacy Leakage in Federated Learning in Radiology Reports: A Comparative Evaluation of Tokenizer-Driven Privacy Risks"},"reference_resolution":{"displayed":49,"state_counts":{"malformed_identifier":3,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":38,"verified_exact":8,"verified_fuzzy":0},"total_outbound_references":49},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-21T06:32:19.484+00:00","source":"crossref"},{"observed_at":"2026-08-21T06:32:16.066871+00:00","source":"retraction_watch"}],"thesis":"As of 22 August 2026, this Paper Citation Record lists 49 of 49 outbound references and 0 inbound Pith citation observations for arXiv:2607.14205."}