{"as_of":"2026-08-15T19:37:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:6c44d0d92d3b2839c116d230d2f1625c2538b954d5e092050aea4750723b29eb","coverage":[{"denominator":10,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":10,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-06T15:29:28.738932Z","state":"measured"},{"denominator":11,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":11,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-15T06:32:42.880941+00:00","state":"measured"},{"denominator":1,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":1,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-02T11:15:48.105239Z","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":[{"citation":{"cited_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2507.15836","snapshot_observed_at":"2026-08-02T11:15:48.105239Z","title":"Optimizing canaries for privacy auditing with metagradient descent.arXiv preprint arXiv:2507.15836, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2606.16952","last_updated":"2026-07-20T23:44:56Z","snapshot_observed_at":"2026-08-14T21:37:37.090182Z","submitted_at":"2026-06-15T16:54:02Z","title":"Phantoms and Disclosures: A Statistical Framework for Auditing Privacy in Synthetic Data","version":2},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-02T11:15:48.105239Z"},"links":{"cited_paper":"/paper/2507.15836","citing_paper":"/paper/2606.16952"},"observation_digest":"sha256:d1eacc6913835ffcaac9871ddb1c496c1c4356e8829252cede2877e2d608cf91","observation_id":"e9cf0070-7f81-46f4-843c-fe04624f8457","resolution":{"observed_at":"2026-08-02T11:15:48.105239Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"links":{"evidence":"/evidence","html":"/paper/2507.15836/citation-record","integrity":"/paper/2507.15836/integrity","json":"/paper/2507.15836/citation-record.json","paper":"/paper/2507.15836"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:29:29.084751Z","title":"Deep learning with differential privacy","venue":null,"work_id":"6c4393ee-ef81-49ab-86c5-f35d39aacf85","year":2016},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.706889Z"},"links":{"citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:2f1593748b869a38b7f4e4fab8d5e8569bd0448f3c08d970164b0a993c4b852c","observation_id":"b602389f-9112-4be8-90fd-936e484f9513","resolution":{"observed_at":"2026-08-06T15:29:29.088745Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-15T06:32:42.880941+00:00","source":"crossref"},{"observed_at":"2026-08-15T06:32:39.529945+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":"2026-08-06T15:29:29.073698Z","title":"Stabilizing differentiable architecture search via perturbation- based regularization","venue":null,"work_id":"a35bf193-74cb-4652-8944-b931dae6c251","year":2020},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.713825Z"},"links":{"citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:86239a0b4c362fc8f8ff54773295c47461844b71cad1ac8b931996e91b5eba72","observation_id":"7032a4e7-3775-4b01-96cb-e272d7c9a12c","resolution":{"observed_at":"2026-08-06T15:29:29.077671Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-15T06:32:42.880941+00:00","source":"crossref"},{"observed_at":"2026-08-15T06:32:39.529945+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":"2026-08-06T15:29:29.063364Z","title":"Tight auditing of differentially private machine learning","venue":null,"work_id":"7c8cde47-7e60-4432-8e71-42247f9e39c1","year":2023},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.728564Z"},"links":{"citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:62f80b7cb2b0d886e51c1a0b91270e7c5d4fe732ce42217cfae43f51dc5344ab","observation_id":"0a691785-1193-4fda-8188-6bb5e43365d2","resolution":{"observed_at":"2026-08-06T15:29:29.066893Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-15T06:32:42.880941+00:00","source":"crossref"},{"observed_at":"2026-08-15T06:32:39.529945+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":"2026-08-06T15:29:29.051778Z","title":"Adversary instantiation: Lower bounds for differentially private machine learning","venue":null,"work_id":"da15556b-c958-42c4-9930-868604377ec1","year":2021},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.732026Z"},"links":{"citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:f1403bac04ccfc79b3dc240d4c32aac641fd50064d21c6b1dc3ee3cb9b0043ec","observation_id":"eec48a3f-a3de-4567-be3f-031c0450e9bf","resolution":{"observed_at":"2026-08-06T15:29:29.056156Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-15T06:32:42.880941+00:00","source":"crossref"},{"observed_at":"2026-08-15T06:32:39.529945+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":"2026-08-06T15:29:29.038464Z","title":"Meta-learning with implicit gradients","venue":null,"work_id":"ec1ba073-3fbd-4949-9d76-1ce79fe847f5","year":2019},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.735284Z"},"links":{"citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:5ccbba38447f5c5cccab8fd819224e65b0364b5b2c92f0661eff6391f59643f5","observation_id":"ef1cda07-1e42-4bbe-a839-f154d62acca1","resolution":{"observed_at":"2026-08-06T15:29:29.044630Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-15T06:32:42.880941+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-15T06:32:42.880941+00:00","source":"crossref"},{"observed_at":"2026-08-15T06:32:39.529945+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":"2026-08-06T15:29:28.720329Z","title":"Forward and reverse gradient-based hyperparameter optimization","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":1135,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.720329Z"},"links":{"citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:7a08068d342fc4798bed2eeb35548991eed375de20b81bb3bac9bbb3a62f1c99","observation_id":"4fe010f6-33d6-48e1-9f52-bc4aa994893c","resolution":{"observed_at":"2026-08-06T15:29:28.720329Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2204.13650","last_updated":"2022-06-16T17:47:42Z","snapshot_observed_at":"2026-08-14T02:25:49.984432Z","submitted_at":"2022-04-28T17:10:56Z","title":"Unlocking High-Accuracy Differentially Private Image Classification through Scale","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2204.13650","snapshot_observed_at":"2026-08-06T15:29:28.717014Z","title":"Unlocking high-accuracy differentially private image classification through scale","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":1565,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.717014Z"},"links":{"cited_paper":"/paper/2204.13650","citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:ad6319c906d5c4b27cd03e7e5d7e92126ed5d163bcb2f27d7eb14633f891b185","observation_id":"b05eb3d5-bd76-40df-a0df-b8a017f1cfaf","resolution":{"observed_at":"2026-08-06T15:29:28.717014Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2202.12219","last_updated":"2022-03-28T17:10:52Z","snapshot_observed_at":"2026-08-13T16:34:38.939518Z","submitted_at":"2022-02-24T17:31:08Z","title":"Debugging Differential Privacy: A Case Study for Privacy Auditing","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2202.12219","snapshot_observed_at":"2026-08-06T15:29:28.738932Z","title":"Wide residual networks","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":2022,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.738932Z"},"links":{"cited_paper":"/paper/2202.12219","citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:585c50e07e590591131639ef26a3a66b83fa3fb0f785cfed6f1b190dfbc6f3e2","observation_id":"b4f19330-916e-4e04-8213-b2882dff0a6c","resolution":{"observed_at":"2026-08-06T15:29:28.738932Z","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":"2026-08-06T15:29:28.724806Z","title":"It’s Our Loss: No Privacy Amplification for Hidden State DP-SGD With Non-Convex Loss","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":2023,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.724806Z"},"links":{"citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:de926e976da2b969e44ae5a56cacc35919e28e89d87bb7f7d3175cc1394cfc25","observation_id":"d17410b7-4dab-4440-b8f9-814a336b687f","resolution":{"observed_at":"2026-08-06T15:29:28.724806Z","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":"2026-08-06T15:29:28.710418Z","title":"Gradient-based optimization of hyperparameters","venue":null,"work_id":null,"year":2000},"citing_paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent","version":1},"reference_index":2025,"source":"pdf_text","source_observed_at":"2026-08-06T15:29:28.710418Z"},"links":{"citing_paper":"/paper/2507.15836"},"observation_digest":"sha256:89e71c7d031bdb4cee31e60a0d05ca84031fb03267a2727aa3a47b0d91bf3034","observation_id":"114a5357-67f0-428f-9ee0-59373ce2d18b","resolution":{"observed_at":"2026-08-06T15:29:28.710418Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2507.15836","last_updated":"2025-07-21T17:47:33Z","latest_version":1,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-10T22:39:07.863214Z","submitted_at":"2025-07-21T17:47:33Z","title":"Optimizing Canaries for Privacy Auditing with Metagradient Descent"},"reference_resolution":{"displayed":10,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":5,"verified_exact":0,"verified_fuzzy":5},"total_outbound_references":10},"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":"2026-08-15T06:32:42.880941+00:00","source":"crossref"},{"observed_at":"2026-08-15T06:32:39.529945+00:00","source":"retraction_watch"}],"thesis":"As of 15 August 2026, this Paper Citation Record lists 10 of 10 outbound references and 1 inbound Pith citation observation for arXiv:2507.15836."}