{"as_of":"2026-08-09T16:06:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:0eb6f49d242020952fec9f0dcb182d7261030942053cbba35be5038de3495b42","coverage":[{"denominator":54,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":54,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-08T23:53:24.523040Z","state":"measured"},{"denominator":54,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":54,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-09T06:31:02.800959+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/2502.04045/citation-record","integrity":"/paper/2502.04045/integrity","json":"/paper/2502.04045/citation-record.json","paper":"/paper/2502.04045"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:23.636625Z","title":"The algorithmic foundations of differential privacy,","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.636625Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:9492ddea2d678fa6b9539fc8b34105562271c5a385f9654b0887a2a42ccf9fe7","observation_id":"992554aa-15d2-4964-a3f6-30cd50eb26ff","resolution":{"observed_at":"2026-08-08T23:53:23.636625Z","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":"2026-08-08T23:53:23.663833Z","title":"Broadening the scope of differential privacy using metrics,","venue":null,"work_id":null,"year":2013},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.663833Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:4ce9d3f8d0b36e8271f6e9d09cd0f8d26202ec6106cd14948029b316afaafe1a","observation_id":"b99893d6-8ac2-4b4a-af71-9e546c96f055","resolution":{"observed_at":"2026-08-08T23:53:23.663833Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.747292Z","title":"Stochastic gradient descent with differentially private updates,","venue":null,"work_id":"95849c00-f3d0-4520-8a4d-fc5ed07518c0","year":2013},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.756574Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:d1b45fc0c7dc1956eb8603fbd30eeef5e2a66789253094e160a018c2820b0550","observation_id":"6ddab337-f5e5-4433-99a5-7b80e8cb0d5a","resolution":{"observed_at":"2026-08-08T23:53:26.751728Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-08T23:53:23.787546Z","title":"Deep learning with differential privacy,","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.787546Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:526b9b2e44b3bd18fdbb6a259c20ffd6bb834b3bb8c1e4ccc17db1ef9518f66e","observation_id":"ff1324bf-c3ad-45a6-a13d-42e07e2e1ac7","resolution":{"observed_at":"2026-08-08T23:53:23.787546Z","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":"2026-08-08T23:53:23.793566Z","title":"Back to the drawing board: Revisiting the design of optimal location privacy-preserving mechanisms,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.793566Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:af6306bce0dd2b480135d9f61702e9c35888bcd5cb89c6c4375aa79f2a81abf6","observation_id":"7223cf17-5b59-41e0-9dba-cf5740f68011","resolution":{"observed_at":"2026-08-08T23:53:23.793566Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.733391Z","title":"Privic: A privacy-preserving method for incremental collection of location data,","venue":null,"work_id":"c54deeaa-fc9f-4742-b9f8-bef610e5252f","year":2024},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.799796Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:5a033cfbb86021a7a877d5ef00d89e4955ea92c1a327e67a4425444adfae6cd8","observation_id":"39d0bef5-09db-477f-87ef-43b137674197","resolution":{"observed_at":"2026-08-08T23:53:26.738223Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.720000Z","title":"A privacy- preserving querying mechanism with high utility for electric vehicles,","venue":null,"work_id":"e0d2a3a6-b379-443a-b463-20f7629a5d1c","year":2024},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.804578Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:a8cc0bd336a16ca9be98e71c7de6d1cd9b2be4c552048f8dd76301611ee179f9","observation_id":"ed0a76c5-12ec-4b84-aa7a-7b7842a0617e","resolution":{"observed_at":"2026-08-08T23:53:26.724453Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.705448Z","title":"Differentially private obfuscation of facial images,","venue":null,"work_id":"c78390a3-fc20-43e0-a293-390eb0c142f4","year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.809151Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:118d74474eec2125dd4f101ea3d7cc7ebe79c1682d2e4f61e6f2187a426ab08b","observation_id":"13bcfbf2-eb58-4bba-b6c1-0e64ffeb1998","resolution":{"observed_at":"2026-08-08T23:53:26.710481Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.507295Z","title":"Differentially private facial obfuscation via generative adversarial networks,","venue":null,"work_id":"a3f57d1f-b7df-4ba7-9fa3-8b5430bf1637","year":2022},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.813408Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:7570a137c4c75f2d9a4e397097172d1f9e0ee67c1f3546400b43fa5bc985f357","observation_id":"e2040043-f88f-44f7-b593-23d905da1190","resolution":{"observed_at":"2026-08-08T23:53:26.611981Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-08T23:53:23.818396Z","title":"Generalised differential privacy for text document processing,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.818396Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:0b3b7e9abdb797609120d7feaf724bda96b808d651249506d4bf84764c48c9ca","observation_id":"73e3175f-f742-4b2a-ac20-ff951e1d6666","resolution":{"observed_at":"2026-08-08T23:53:23.818396Z","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":"2026-08-08T23:53:23.822824Z","title":"Leveraging hierarchical representations for preserving privacy and utility in text,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.822824Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:36cdc8586d5fa92fa00f6129a9a15d02218ba8e4eb55ba4f4f87960cee3f9355","observation_id":"a336340d-3316-4f34-9af1-a52802742b6e","resolution":{"observed_at":"2026-08-08T23:53:23.822824Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.448341Z","title":"Differential privacy in natural language processing the story so far,","venue":null,"work_id":"564d0ebd-c41e-4414-a63c-caa7732c8d9f","year":2022},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.827518Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:a64befcb44db60b15bf315b1663c6d91c58fbfd8510580198e0a72abaa5fe26b","observation_id":"c648c490-fa66-4b47-846c-ab98ee706c71","resolution":{"observed_at":"2026-08-08T23:53:26.453160Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.432220Z","title":"Group privacy for personalized federated learning,","venue":null,"work_id":"c21424a4-a0e3-44bd-b6d0-b787be87d5bc","year":2022},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.831883Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:4baea627de2f2cb5359eb59fa38b7e6825bc6e59019df89f19cbbd211eab493e","observation_id":"796a35df-d708-48f7-bda4-bb0f1d69be17","resolution":{"observed_at":"2026-08-08T23:53:26.438312Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.415587Z","title":"Ad- vancing personalized federated learning: Group privacy, fairness, and beyond,","venue":null,"work_id":"1500d244-43ac-487a-abc6-efee41dd2a0d","year":2023},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.836284Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:143e6b7689e575cf4209598c76e98140962e3ca203147637c648ae4d5c65d208","observation_id":"4b522dc4-f394-41e9-bf40-6830e7308eeb","resolution":{"observed_at":"2026-08-08T23:53:26.421250Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.252576Z","title":"Directional privacy for deep learning,","venue":null,"work_id":"0dd8dcd7-eab2-4cc7-bfed-1197b1f6e1bd","year":2023},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.840440Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:b2820165658ff2656fc5223e1c1b38b5f389a9152bfdd23a67edf212e5677139","observation_id":"6b6780b9-fa54-43ad-ac62-61887074acac","resolution":{"observed_at":"2026-08-08T23:53:26.344304Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-08T23:53:23.844783Z","title":"Differential privacy for directional data,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.844783Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:0bf9b932d8528f451e5bcb5618350e33aa713ce1bbea0b0575ec2d62073c819c","observation_id":"8b3a7cd0-8cb1-4726-8428-dd1fc7d4ba7d","resolution":{"observed_at":"2026-08-08T23:53:23.844783Z","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":"2026-08-08T23:53:23.849262Z","title":"Exploiting unintended feature leakage in collaborative learning,","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.849262Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:3b51a2191fdfa02dd5b057cc86d97e1965c5da1a92f8bcc21a44e666720f4d78","observation_id":"1d581313-f13f-408b-b2ca-16ef441d2db7","resolution":{"observed_at":"2026-08-08T23:53:23.849262Z","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":"2026-08-08T23:53:23.853245Z","title":"Reconstructing training data with informed adversaries,","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.853245Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:33f7edfb231dba708e5afd3ecb3df3563ca830cedcc258f5bb0fafeba6fa28c3","observation_id":"da39afe7-aea0-47ec-9204-0448c55eb3cf","resolution":{"observed_at":"2026-08-08T23:53:23.853245Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.139091Z","title":"R ´enyi differential privacy,","venue":null,"work_id":"ee863de5-a63e-46ea-be8c-8d04ca44b92d","year":2017},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.857623Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:ee59e6fc37de8ef63bc95ee7fab5c59ec7dac8919a613a7d7afc2d8086e72693","observation_id":"4fde2dc8-d174-4d77-8ee4-0de8fbce9a76","resolution":{"observed_at":"2026-08-08T23:53:26.182461Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.121342Z","title":"Comparing systems: Max-case refinement orders and application to differential privacy,","venue":null,"work_id":"3b1f8a61-1818-4882-ad42-c4eb433148c3","year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.862155Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:9a1ce54c614a1222d4e2fb6e74e9fb89d9a5df81ba641b63eb9910d794c93d62","observation_id":"d607ccdc-7cac-4bc0-96ea-63f9e48f3013","resolution":{"observed_at":"2026-08-08T23:53:26.126930Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.104466Z","title":null,"venue":null,"work_id":"a22d96bf-017e-44a5-88b2-573cb31a3ca8","year":2020},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.866708Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:d76b1165c22ea6620cfcd9bb3a6619a1efd8a75bddb3f46705e86c39ecfcb0fb","observation_id":"cbf67b5d-d8d4-4f56-84d3-a98a71221614","resolution":{"observed_at":"2026-08-08T23:53:26.109782Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-08T23:53:23.871359Z","title":"R ´enyi differential privacy,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.871359Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:f71d853464fc866381e3146f136b1d5b9f35bc14329a013483f9182266ecd64f","observation_id":"7da5306d-9d02-4481-b39a-7031b789009f","resolution":{"observed_at":"2026-08-08T23:53:23.871359Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1908.10530","last_updated":"2019-08-28T03:03:25Z","snapshot_observed_at":"2026-08-07T03:15:48.518041Z","submitted_at":"2019-08-28T03:03:25Z","title":"R\\'enyi Differential Privacy of the Sampled Gaussian Mechanism","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1908.10530","snapshot_observed_at":"2026-08-08T23:53:23.875876Z","title":"R \\’enyi differential privacy of the sampled gaussian mechanism,","venue":null,"work_id":null,"year":1908},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.875876Z"},"links":{"cited_paper":"/paper/1908.10530","citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:098d2ac396302fb44cff65e87bd957c0978e5f35f52fc929bb215b43be9d830c","observation_id":"d68e9632-5281-4552-90b3-b8430dad1af2","resolution":{"observed_at":"2026-08-08T23:53:23.875876Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.080479Z","title":"Poission subsampled r´enyi differential privacy,","venue":null,"work_id":"cbf2dbce-3dfb-4958-a70c-6100250c220b","year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.914060Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:78215b86484cb19b6c546e32a977cc8459eaf892b18574021691d4f219dedc97","observation_id":"c07a0b76-44e1-4706-953a-304ab6c427bf","resolution":{"observed_at":"2026-08-08T23:53:26.084994Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-08T23:53:23.955727Z","title":"Privacy amplification by sub- sampling: Tight analyses via couplings and divergences,","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:23.955727Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:5ec0ca979a3119ca99c293352436fd3be299f9a04a15cc5db716813bde28b579","observation_id":"1bc8e34a-73b1-448b-bb52-45c4d7d9f77d","resolution":{"observed_at":"2026-08-08T23:53:23.955727Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.056937Z","title":"Subsampled r ´enyi differential privacy and analytical moments accountant,","venue":null,"work_id":"4230ef1c-5059-434c-9ee3-765ce3669fb1","year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.003250Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:b70fa2514a5c36f4df87bfda578b7ef24a8a4515da79dbeafb75eb4601d36d52","observation_id":"28b5cf66-33e9-42ec-8bf8-330752544a1d","resolution":{"observed_at":"2026-08-08T23:53:26.061428Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.043007Z","title":"A better bound gives a hundred rounds: Enhanced privacy guarantees via f- divergences,","venue":null,"work_id":"7f4ac153-6cb3-47bf-98a5-47a31b0ebfc5","year":2020},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.070589Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:02e957b771af6601d461eaf88f932952554e4a940048c8748e633ea587dc3104","observation_id":"bcaaee7a-b6e9-4d7c-8107-02c8f622fc23","resolution":{"observed_at":"2026-08-08T23:53:26.047760Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.029549Z","title":"The discrete gaussian for differential privacy,","venue":null,"work_id":"7b12c2c9-86fd-4e41-bdc9-f08b204576b7","year":2022},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.124202Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:f14105a4493f15ee34674b9ed4c2fcbb5c63e3968e4372baf960a821691455da","observation_id":"35a12f8f-b4bc-45e5-9db7-7f6c9b75d392","resolution":{"observed_at":"2026-08-08T23:53:26.034214Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.015390Z","title":"Evaluating differentially private machine learning in practice,","venue":null,"work_id":"fb50c735-8510-4468-9f26-008b24fe0453","year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.158275Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:d5b88af02b915dcf96bb8dc2fccafdb046150e961e981d92a1a61e0344bc1d11","observation_id":"d29b7178-793d-4d75-9ac9-11c64bc0be36","resolution":{"observed_at":"2026-08-08T23:53:26.019889Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:26.001280Z","title":"Deep leakage from gradients,","venue":null,"work_id":"298de7a2-874a-45f3-9c94-0e0bd5022e0c","year":2019},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.190531Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:b82b53df774e68ece22f99dabc8e01a4d9b00f1eb3d32adfbae96204b118ebdb","observation_id":"38c756df-dd40-4a5d-9651-f40378e43438","resolution":{"observed_at":"2026-08-08T23:53:26.006060Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2001.02610","last_updated":"2020-01-08T16:45:09Z","snapshot_observed_at":"2026-08-06T23:49:37.129499Z","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":"2026-08-08T23:53:24.221834Z","title":"idlg: Improved deep leakage from gradients,","venue":null,"work_id":null,"year":2001},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.221834Z"},"links":{"cited_paper":"/paper/2001.02610","citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:8e7e83921fcaa49de79829cd354b929caf2fc621d074e7655b2876cb5bf266b8","observation_id":"5e7b6d6c-2e8b-444c-8839-948424f204b4","resolution":{"observed_at":"2026-08-08T23:53:24.221834Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.988481Z","title":"Inverting gradients - how easy is it to break privacy in federated learning?","venue":null,"work_id":"c752f271-c709-403a-9a4e-81176757f86d","year":2020},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.285621Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:580ab483f5b3fbc3becef1bf6e7ca6461b4c998bfb808825e1b46a56f9dc1bcd","observation_id":"f185c1a6-6527-4956-a2a1-619295f521eb","resolution":{"observed_at":"2026-08-08T23:53:25.992872Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.974885Z","title":"Evaluating gradient inversion attacks and defenses in federated learning,","venue":null,"work_id":"54b1eeda-0761-47a5-ab1d-b01698a1f1f3","year":2021},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.344999Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:599eec73245a1dc9bcee6bd63543edafab6d016af32faf1b14e3ea1a8aa5a5d9","observation_id":"016a20f3-7f5a-438a-a50a-51aa02a136e9","resolution":{"observed_at":"2026-08-08T23:53:25.979406Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.960042Z","title":"Learning to invert: Simple adaptive attacks for gradient inversion in federated learning,","venue":null,"work_id":"7a126972-6f28-4091-97c4-a3061c6f4031","year":2023},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.407486Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:392f3cf0c8467fa8af74dc36ab1a9a78f219f73e7dc611aeedb712200081fbeb","observation_id":"9330d47b-edee-4787-b3ad-2c25ba59340a","resolution":{"observed_at":"2026-08-08T23:53:25.964977Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.743096Z","title":"Reconstructing training data from model gradient, provably,","venue":null,"work_id":"8c6e084e-cfa1-4361-8721-69a4655989d4","year":2023},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.438406Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:9d893a84c2742458a5430911ea8227c08ba17aa4bf1fc3ff8473b5c4ff75bfa1","observation_id":"65b45036-4e96-4e05-b20f-65a36eeb1d99","resolution":{"observed_at":"2026-08-08T23:53:25.806255Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.616300Z","title":"Sok: Gradient leakage in federated learning,","venue":null,"work_id":"318f7278-c5d0-42e4-99a4-4468bf84ef27","year":2024},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.443018Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:d2f6293db2405ca5a7d04d7e4c56228e1547be44e15e7a02ec330b74d2f8bead","observation_id":"f95ce94d-bb60-4d60-8b64-ea41a4054c30","resolution":{"observed_at":"2026-08-08T23:53:25.645337Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.508348Z","title":"On the Foundations of Quantitative Information Flow,","venue":null,"work_id":"2ec98d85-ecaa-4be3-85f9-57a738f6e1eb","year":2009},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.447297Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:4b8155dc76178793da5fd46aff9cd17d7f4ca66c4512f458487f0a844947eeba","observation_id":"a4f5326a-ba64-4345-860a-31c0eadad03a","resolution":{"observed_at":"2026-08-08T23:53:25.559642Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.493614Z","title":"An operational approach to information leakage,","venue":null,"work_id":"68d97d21-840c-4f7c-a765-a79a2b27e859","year":2020},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.451528Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:5b864cc66afc58d659807ad33763fc3cdcb8779b7513783111fe35d31c05827d","observation_id":"0e379a86-f145-48fe-b759-7387243e548f","resolution":{"observed_at":"2026-08-08T23:53:25.497696Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.480109Z","title":"Information radius,","venue":null,"work_id":"c420b037-577f-4ff6-9256-550c44808138","year":1969},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.456255Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:fa63359256116bf7ea7c6c2687f83121cb7287d626a744a8e59670f30a2ac803","observation_id":"d5257238-0d9c-4504-a9a3-05a19a45f790","resolution":{"observed_at":"2026-08-08T23:53:25.484659Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2210.12916","last_updated":"2023-05-18T07:22:56Z","snapshot_observed_at":"2026-08-03T13:10:44.205405Z","submitted_at":"2022-10-24T02:03:22Z","title":"Explaining epsilon in local differential privacy through the lens of quantitative information flow","version":2},"cited_work":{"arxiv_id":"2210.12916","doi":null,"metadata_source":"pith","pith_arxiv_id":"2210.12916","snapshot_observed_at":"2026-08-08T23:53:24.813076Z","title":"Explaining epsilon in local differential privacy through the lens of quantitative information flow","venue":"cs.IT","work_id":"a0a07b14-65c9-4be1-817d-413ada6e3963","year":2022},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.460920Z"},"links":{"cited_paper":"/paper/2210.12916","citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:76dd8381dfe23f3f610ceba681ff8065c8c9f5c60f344bf2bfdbae651999ba85","observation_id":"3a83f1ff-a0dd-47ed-921b-47e137bafee3","resolution":{"observed_at":"2026-08-08T23:53:24.818181Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.464877Z","title":"Closed-form bounds for dp-sgd against record-level inference attacks","venue":null,"work_id":"822692d1-dd31-4fa3-be4d-4ce65574cbc3","year":null},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.466187Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:efe9b803fc3da77a3cf8256e152b286db40ce3b2245a12202434671a1f9bd1d0","observation_id":"7255a543-bbf7-4198-9dc0-597503d00695","resolution":{"observed_at":"2026-08-08T23:53:25.470358Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.450740Z","title":"Bounding training data reconstruction in private (deep) learning,","venue":null,"work_id":"4d5f14d2-26fc-4071-8073-e002e8335eb6","year":2022},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.471187Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:e5f6ad324ab621fe4c5f8282b48ad99e06bfcd1eac2b2a91c752c172e2c6bb07","observation_id":"5b28b7c9-4090-4902-ae9b-896301b3d277","resolution":{"observed_at":"2026-08-08T23:53:25.455331Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.437245Z","title":"Bounding training data reconstruction in DP-SGD,","venue":null,"work_id":"18422489-3f06-4fa3-b8fc-8390cddffc67","year":2023},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.475811Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:6115950640c9b501b735e33e1ac569194fa9763bb09a66cd81ede061e61fc958","observation_id":"957b5990-6fb6-46f0-9bfa-322e36fde14f","resolution":{"observed_at":"2026-08-08T23:53:25.441544Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-08T23:53:24.480421Z","title":"Calibrating noise to sensitivity in private data analysis,","venue":null,"work_id":null,"year":2006},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.480421Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:05d13bdd0e6537e86f2edcd63336c7e38f27476e2f84854c9a05f071d5e9e2f9","observation_id":"de223f3c-3429-4052-bd97-1740e3e9b87a","resolution":{"observed_at":"2026-08-08T23:53:24.480421Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2202.05192","last_updated":"2022-11-19T13:35:12Z","snapshot_observed_at":"2026-08-09T14:33:03.747421Z","submitted_at":"2022-02-10T17:50:21Z","title":"von Mises-Fisher distributions and their statistical divergence","version":2},"cited_work":{"arxiv_id":"2202.05192","doi":null,"metadata_source":"pith","pith_arxiv_id":"2202.05192","snapshot_observed_at":"2026-08-08T23:53:24.746593Z","title":"von Mises-Fisher distributions and their statistical divergence","venue":"econ.EM","work_id":"459f614b-a59f-4fa0-9d34-a6fe96dee190","year":2022},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.484891Z"},"links":{"cited_paper":"/paper/2202.05192","citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:254d1f4af038907022b3a271f02a150fdc8ae43c0c55ac3cce6277b95a4a69df","observation_id":"1c970636-c48f-4229-a57e-0ace27157af9","resolution":{"observed_at":"2026-08-08T23:53:24.774431Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-08T23:53:24.489499Z","title":"The composition theorem for differential privacy,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.489499Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:7ed0d15d61112b08ca59b06d16fba3cf43f762c19efb517093236cdb48695225","observation_id":"bf50fde0-4a99-48f7-afce-01166acdab07","resolution":{"observed_at":"2026-08-08T23:53:24.489499Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.13569","last_updated":"2024-06-19T13:58:42Z","snapshot_observed_at":"2026-08-09T14:34:45.509598Z","submitted_at":"2024-06-19T13:58:42Z","title":"Bayes' capacity as a measure for reconstruction attacks in federated learning","version":1},"cited_work":{"arxiv_id":"2406.13569","doi":null,"metadata_source":"pith","pith_arxiv_id":"2406.13569","snapshot_observed_at":"2026-08-08T23:53:24.627269Z","title":"Bayes' capacity as a measure for reconstruction attacks in federated learning","venue":"cs.LG","work_id":"9c549744-cd54-47c1-b21b-1c4189ae0b8d","year":2024},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.493476Z"},"links":{"cited_paper":"/paper/2406.13569","citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:20693fdc07a5623072badedd94d3a80d0a4ef185e26442d5b71496cf32c40eb6","observation_id":"c536e60e-47fc-474f-b20e-2b2d71b29cd9","resolution":{"observed_at":"2026-08-08T23:53:24.700326Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.403223Z","title":"See through gradients: Image batch recovery via gradinversion,","venue":null,"work_id":"cf8c0932-9f24-4920-9b14-51c1e1dff6b8","year":2021},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.498174Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:f8a66bfd250f21183952fab4d90516902786a5a3d70af61d8f9cc96a6ecff088","observation_id":"4564cd27-cff6-4e0d-8f96-d2ccbb273c9d","resolution":{"observed_at":"2026-08-08T23:53:25.408031Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.387975Z","title":"Mea- suring information leakage using generalized gain functions,","venue":null,"work_id":"e21170c9-95d3-48dd-b8c7-db44ec943e4f","year":2012},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.502023Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:e3440f15095c849aed76bb9be89dbf9bff741b5cd173c6391d492c5605cd7572","observation_id":"bf3e85c6-fef7-4091-84da-ee579bb66866","resolution":{"observed_at":"2026-08-08T23:53:25.392424Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+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":"2026-08-08T23:53:24.506094Z","title":"The mnist database of handwritten digit images for machine learning research,","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.506094Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:c5f3ac6047c73d56e29b99632e2363638f321c9f5a60b96457bf4fed8312eea1","observation_id":"47615073-3c6e-4e9f-937a-6d9e6eb80ca0","resolution":{"observed_at":"2026-08-08T23:53:24.506094Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1708.07747","last_updated":"2017-09-15T21:29:49Z","snapshot_observed_at":"2026-07-06T05:56:41.814255Z","submitted_at":"2017-08-25T14:01:29Z","title":"Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1708.07747","snapshot_observed_at":"2026-08-08T23:53:24.509975Z","title":"Fashion-mnist: a novel image dataset for benchmarking machine learning algorithms,","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.509975Z"},"links":{"cited_paper":"/paper/1708.07747","citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:eeefb34a57b6c3bbaa5fea2eb20154e83ed50e93bbb479ae26bd660c4830be90","observation_id":"3049c233-00ce-442b-a99a-fc1d3dde9c14","resolution":{"observed_at":"2026-08-08T23:53:24.509975Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2109.12298","last_updated":"2022-08-22T20:24:50Z","snapshot_observed_at":"2026-08-06T15:46:33.397187Z","submitted_at":"2021-09-25T07:10:54Z","title":"Opacus: User-Friendly Differential Privacy Library in PyTorch","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2109.12298","snapshot_observed_at":"2026-08-08T23:53:24.514234Z","title":"Opacus: User-friendly differential privacy library in pytorch,","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.514234Z"},"links":{"cited_paper":"/paper/2109.12298","citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:636f45d60325ac268b1ab1b9403b79ca9f72d037455d3fca403441dd1c571859","observation_id":"4f0c9b1b-fd62-4b3b-ae0f-2370c50b9bea","resolution":{"observed_at":"2026-08-08T23:53:24.514234Z","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":"2026-08-08T23:53:24.518425Z","title":"Image quality assess- ment: from error visibility to structural similarity,","venue":null,"work_id":null,"year":2004},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.518425Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:5875ea185afbd7361de0876960a600b68e28c2f30ca99198199e882d04b67235","observation_id":"905e0961-4410-464e-8999-03ebecb332d0","resolution":{"observed_at":"2026-08-08T23:53:24.518425Z","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":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-08T23:53:25.355374Z","title":"Privacy assessment on reconstructed images: Are existing evaluation metrics faithful to human perception?","venue":null,"work_id":"3e142dda-7b68-4c30-8b5d-66dc4aacb7f8","year":2023},"citing_paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning","version":1},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-08-08T23:53:24.523040Z"},"links":{"citing_paper":"/paper/2502.04045"},"observation_digest":"sha256:f0bc7b225fe38ed4dc45f1843cb49f0b66a01df41e15f125493edc65691c201e","observation_id":"c5313740-3b8d-40c3-8235-b4e401483cf7","resolution":{"observed_at":"2026-08-08T23:53:25.360384Z","resolver_source":"raw_fallback","status":"malformed_identifier"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2502.04045","last_updated":"2025-02-06T13:04:25Z","latest_version":1,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-09T14:34:50.620963Z","submitted_at":"2025-02-06T13:04:25Z","title":"Comparing privacy notions for protection against reconstruction attacks in machine learning"},"reference_resolution":{"displayed":54,"state_counts":{"malformed_identifier":2,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":19,"verified_exact":3,"verified_fuzzy":30},"total_outbound_references":54},"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":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"thesis":"As of 9 August 2026, this Paper Citation Record lists 54 of 54 outbound references and 0 inbound Pith citation observations for arXiv:2502.04045."}