{"as_of":"2026-08-18T18:23:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:f5349bd51282febe62fee5bcbb8029c625c9b3a1b4f4aa3d47d8edc8a3c01f64","coverage":[{"denominator":55,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":55,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-04T19:11:57.458158Z","state":"measured"},{"denominator":55,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":55,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-18T06:34:40.430872+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/2509.09485/citation-record","integrity":"/paper/2509.09485/integrity","json":"/paper/2509.09485/citation-record.json","paper":"/paper/2509.09485"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-04T19:11:47.033221Z","title":"Deep learning with differential privacy","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":1,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.033221Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:d8c54437a919bdff293aee71710190e684c88eac178ee17c5b9fb4f5adac980c","observation_id":"c08c835c-1528-4ccd-a538-69144eaf95ea","resolution":{"observed_at":"2026-08-04T19:11:47.033221Z","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":"2026-08-04T19:11:47.098941Z","title":"Database-friendly random projections","venue":null,"work_id":null,"year":2001},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":2,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.098941Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:daddd74565690355a82628839aa1d3b0bfe5e3de83a86a3590b880830728729b","observation_id":"4aa2e6ea-341e-4d35-9e16-533c00d69789","resolution":{"observed_at":"2026-08-04T19:11:47.098941Z","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":"2026-08-04T19:11:47.156458Z","title":"cpsgd: Communication-efficient and differentially-private distributed sgd","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":3,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.156458Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:c5f023df9c129da9d1f0c75b4c6ff177fe5c739e34b458b2b2b9f0446354ead0","observation_id":"f9a1356c-0e43-49e0-9f6e-f16f5442c0c1","resolution":{"observed_at":"2026-08-04T19:11:47.156458Z","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":"2026-08-04T19:11:47.265393Z","title":"Private empirical risk minimization: Efficient algorithms and tight error bounds","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":4,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.265393Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:fdef3e131f524f141be814a63661ffa29f948322736dca236a19d16bc8b144de","observation_id":"ff8568d9-a85b-4966-a975-3d09929ceb26","resolution":{"observed_at":"2026-08-04T19:11:47.265393Z","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":"2026-08-04T19:11:47.372591Z","title":"A critical review on the use (and misuse) of differential privacy in machine learning","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":5,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.372591Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:b18fd651253f4f7ada82f54d9d98d8d3cd2c348549562402cc4ad4bb9ff5eb60","observation_id":"f9943d3c-ddf1-46d9-aa60-ca4d9b398b88","resolution":{"observed_at":"2026-08-04T19:11:47.372591Z","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":"2026-08-04T19:11:47.464038Z","title":"The johnson-lindenstrauss transform itself preserves differential privacy","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":6,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.464038Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:7ea8e7a7579a6c34059713035b6d61325342788adce8bed7fac71b577017f527","observation_id":"82cf2b23-4c81-4e28-9e02-1ea338bfd6a6","resolution":{"observed_at":"2026-08-04T19:11:47.464038Z","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":"2026-08-04T19:11:47.565736Z","title":"Optimization methods for large-scale machine learning","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":7,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.565736Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:c8229d0609223ceb98b9fb2121c74be50529b674c15bf408efe80aa2dfa015e0","observation_id":"5bc1f99e-f4aa-4f73-8587-685e09425832","resolution":{"observed_at":"2026-08-04T19:11:47.565736Z","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":"2026-08-04T19:11:47.659111Z","title":"Automatic clipping: Differentially private deep learning made easier and stronger","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":8,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.659111Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:c04ee8e8fba141c89303e2426abf8916effeb80752de175e33ff5a8322d6cdd5","observation_id":"61c083d1-7184-4f4c-81da-4826790e633f","resolution":{"observed_at":"2026-08-04T19:11:47.659111Z","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":"2026-08-04T19:11:47.759130Z","title":"Model compression","venue":null,"work_id":null,"year":2006},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":9,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.759130Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:618a7cad78ce59a2b505d73001e3cff547eba0297b53428086f31933b3826325","observation_id":"23f8fe3e-2317-4cc8-b2a3-1b2d3505e72b","resolution":{"observed_at":"2026-08-04T19:11:47.759130Z","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":"2026-08-04T19:11:47.851465Z","title":"Ethical machine learning in healthcare","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":10,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.851465Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:9395dca8e590147f6d8f595aba61010205106bb6247c12d8dfbf010b30e61233","observation_id":"467b5e5a-7e26-444d-8052-573d31370670","resolution":{"observed_at":"2026-08-04T19:11:47.851465Z","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":"2026-08-04T19:11:47.921476Z","title":"Understanding gradient clipping in private sgd: A geometric perspective","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":11,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:47.921476Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:6a33fee027f55bb09e59a87d9f39c271735843b9b888ae48e7c1b44997ab1907","observation_id":"892581b2-806d-44c1-9344-6b146bf26531","resolution":{"observed_at":"2026-08-04T19:11:47.921476Z","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":"2026-08-04T19:11:48.015212Z","title":"Robust quantization: One model to rule them all","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":12,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.015212Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:454b1f2e5cd73be066144d7d046b6db934b760feddffae95c803166de0a458df","observation_id":"83114b9f-a79a-4cc8-bce7-4d92d888eff8","resolution":{"observed_at":"2026-08-04T19:11:48.015212Z","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":"2026-08-04T19:11:48.086670Z","title":"A comprehensive survey on model compression and acceleration","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":13,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.086670Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:9521c1f3660abb06ce7ca9b41a044685a5550175faa68a981b3b3847a5f78b58","observation_id":"988cf950-58f7-419d-8f84-ec2edef005b2","resolution":{"observed_at":"2026-08-04T19:11:48.086670Z","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":"2026-08-04T19:11:48.148224Z","title":"Beyond uniform lipschitz condition in differentially private optimization","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":14,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.148224Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:5b10abc7f8aea49fb6a3d26620b1f2b09e7bb1bad9a4feb11e354c593786bb16","observation_id":"6780556e-e2cc-4182-8ad6-87a243c594bc","resolution":{"observed_at":"2026-08-04T19:11:48.148224Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1206.6813","last_updated":"2012-06-27T15:36:47Z","snapshot_observed_at":"2026-08-15T00:55:15.174632Z","submitted_at":"2012-06-27T15:36:47Z","title":"A concentration theorem for projections","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1206.6813","snapshot_observed_at":"2026-08-04T19:11:48.220130Z","title":"A concentration theorem for projections","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":15,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.220130Z"},"links":{"cited_paper":"/paper/1206.6813","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:c9efffa7f0a295ab3ba28f7e8a21e20438eb9cd58f83f21c9612d38f82d01089","observation_id":"ada54b33-f2f7-4d36-b96a-7c500bb057f4","resolution":{"observed_at":"2026-08-04T19:11:48.220130Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2111.00173","last_updated":"2022-01-17T21:31:36Z","snapshot_observed_at":"2026-08-16T17:45:16.546794Z","submitted_at":"2021-10-30T04:45:11Z","title":"Dynamic Differential-Privacy Preserving SGD","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2111.00173","snapshot_observed_at":"2026-08-04T19:11:48.309059Z","title":"Dynamic differential-privacy preserving sgd","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":16,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.309059Z"},"links":{"cited_paper":"/paper/2111.00173","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:683640dc3c7770b261bb02262b95b7b9c9e31cb8e81ca5113cce9df6429e0cc3","observation_id":"60244d8b-4800-4f20-8196-b2b9ddeb5475","resolution":{"observed_at":"2026-08-04T19:11:48.309059Z","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":"2026-08-04T19:11:48.383616Z","title":"Differential privacy","venue":null,"work_id":null,"year":2006},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":17,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.383616Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:6570368cf1eae1caec3ccb88ebc50bc289590fd0f6c06984d19f0a4e56f398af","observation_id":"2ad80216-79a1-444e-a7a1-73be786e6f1a","resolution":{"observed_at":"2026-08-04T19:11:48.383616Z","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":"2026-08-04T19:11:48.457546Z","title":"Differential privacy: A survey of results","venue":null,"work_id":null,"year":2008},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":18,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.457546Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:ab25af19c2444b221d59965af0a4e9761ecb95f5b6adf6e3f44bf9c31be55c8b","observation_id":"0c830a59-d499-4d59-8ef2-4331b7b490fe","resolution":{"observed_at":"2026-08-04T19:11:48.457546Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.06606","last_updated":"2024-02-09T18:34:08Z","snapshot_observed_at":"2026-08-16T14:19:47.882102Z","submitted_at":"2024-02-09T18:34:08Z","title":"RQP-SGD: Differential Private Machine Learning through Noisy SGD and Randomized Quantization","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.06606","snapshot_observed_at":"2026-08-04T19:11:48.539565Z","title":"Rqp-sgd: Differential private machine learning through noisy sgd and randomized quantization","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":19,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.539565Z"},"links":{"cited_paper":"/paper/2402.06606","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:673242eae3706c7d1664107fd777c23d6ecd244d9e4d2f50e57e428f8f2055a4","observation_id":"cc28d259-be2f-4af7-96b4-32c9baa9c471","resolution":{"observed_at":"2026-08-04T19:11:48.539565Z","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":"2026-08-04T19:11:48.606263Z","title":"Semi-supervised learning using deep generative models and auxiliary tasks","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":20,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.606263Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:1cb4836a5e8f8d4eb02eda472f462a53d0ef6afd0fdf19a195ab9929464f25cc","observation_id":"8a704470-c79a-4cf6-8b52-bf4b1170716b","resolution":{"observed_at":"2026-08-04T19:11:48.606263Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2403.15638","last_updated":"2024-04-26T20:24:11Z","snapshot_observed_at":"2026-08-17T13:03:25.719184Z","submitted_at":"2024-03-22T22:27:44Z","title":"Differentially Private Next-Token Prediction of Large Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.15638","snapshot_observed_at":"2026-08-04T19:11:48.680282Z","title":"Differentially private next-token prediction of large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":21,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.680282Z"},"links":{"cited_paper":"/paper/2403.15638","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:438321fd9f84b6b2d44d9ec99ff246e94d321080cc9b8a4460eeba0787c46d02","observation_id":"ab759da2-91cb-41d4-beb0-ec271609a92d","resolution":{"observed_at":"2026-08-04T19:11:48.680282Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2301.11235","last_updated":"2024-03-09T13:28:29Z","snapshot_observed_at":"2026-08-16T15:59:56.095916Z","submitted_at":"2023-01-26T17:18:36Z","title":"Handbook of Convergence Theorems for (Stochastic) Gradient Methods","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2301.11235","snapshot_observed_at":"2026-08-04T19:11:48.753617Z","title":"Handbook of convergence theorems for (stochastic) gradient methods","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":22,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.753617Z"},"links":{"cited_paper":"/paper/2301.11235","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:b1b756d81f823ec94bb9c0324fba4aca2f355127bbf9c146aaa85ce80389c59f","observation_id":"f2710097-b845-44d4-8a0e-cb8e6ffa8dd0","resolution":{"observed_at":"2026-08-04T19:11:48.753617Z","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":"2026-08-04T19:11:48.842519Z","title":"Stochastic first-and zeroth-order methods for nonconvex stochastic programming","venue":null,"work_id":null,"year":2013},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":23,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.842519Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:a0ee9f0e71b3418c19249d24e1f447ad5e29f0d37b7a63496d154d0c53eef7a1","observation_id":"71c743a2-a4ab-4b5d-a2d0-d91606f2b5fe","resolution":{"observed_at":"2026-08-04T19:11:48.842519Z","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":"2026-08-04T19:11:48.902289Z","title":"Artificial intelligence and machine learning in finance: Identifying foundations, themes, and research clusters from bibliometric analysis","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":24,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:48.902289Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:364d092fee541e943ec51931b3c4e8e9ee1fbeb049116b02ee8063b36014a7c1","observation_id":"a060f7f1-3af6-439e-85fe-f141146b8c47","resolution":{"observed_at":"2026-08-04T19:11:48.902289Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2303.01256","last_updated":"2025-07-23T18:17:16Z","snapshot_observed_at":"2026-08-17T17:14:21.805232Z","submitted_at":"2023-03-02T13:36:28Z","title":"Choosing Public Datasets for Private Machine Learning via Gradient Subspace Distance","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.01256","snapshot_observed_at":"2026-08-04T19:11:56.986690Z","title":"Choosing public datasets for private machine learning via gradient subspace distance","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":25,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:56.986690Z"},"links":{"cited_paper":"/paper/2303.01256","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:504ea6b41f3efc4c9d2aab83e4f0678122aa114767b0ded9ac8f81bf5b039146","observation_id":"d686d55a-ff0e-49bd-a37d-867ad5936a23","resolution":{"observed_at":"2026-08-04T19:11:56.986690Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1412.7584","last_updated":"2014-12-24T01:51:06Z","snapshot_observed_at":"2026-08-14T23:06:13.408354Z","submitted_at":"2014-12-24T01:51:06Z","title":"Differential Privacy and Machine Learning: a Survey and Review","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1412.7584","snapshot_observed_at":"2026-08-04T19:11:57.307742Z","title":"Differential privacy and machine learning: a survey and review","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":26,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.307742Z"},"links":{"cited_paper":"/paper/1412.7584","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:ed280596fe5e2560a5db1047ac32020cd69b91720054191b61ecedb02f79d40d","observation_id":"35aae15a-bb10-4deb-b70c-d5678e6650b9","resolution":{"observed_at":"2026-08-04T19:11:57.307742Z","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":"2026-08-04T19:11:57.313379Z","title":"Extensions of lipschitz mappings into a hilbert space","venue":null,"work_id":null,"year":1984},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":27,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.313379Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:c333550648f71e4114c4f3120ab2f2e4ab32756f45a4edae170e555a01f7c634","observation_id":"3f34ee7d-586f-47ef-b21a-10105bee7195","resolution":{"observed_at":"2026-08-04T19:11:57.313379Z","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":"2026-08-04T19:11:57.318481Z","title":"Sgd with low-dimensional gradients with applications to private and distributed learning","venue":null,"work_id":null,"year":1905},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":28,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.318481Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:74e5fcfc9adad06ab1f64596bbb0aad458493bc75fb713681c73b415fa564aa7","observation_id":"f035d801-2b00-4e87-8c78-dc9c3328e874","resolution":{"observed_at":"2026-08-04T19:11:57.318481Z","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":"2026-08-04T19:11:57.323752Z","title":"u chemann, Maria Bannert, Daryna Dementieva, Frank Fischer, Urs Gasser, Georg Groh, Stephan G \\","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":29,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.323752Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:52887cad10279202d2cbda4422e608625a76092fa720242acef9d9ced7bbad21","observation_id":"2057a9b3-9c05-435b-8514-3e81527c9921","resolution":{"observed_at":"2026-08-04T19:11:57.323752Z","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":"2026-08-04T19:11:57.328830Z","title":"Revisiting gradient clipping: Stochastic bias and tight convergence guarantees","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":30,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.328830Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:ef0441093b08a3b792a9cec58f0dfd9b4da12ef6b785138147fe579a44772074","observation_id":"301bb459-5a2d-4e28-8f38-563a1a55472d","resolution":{"observed_at":"2026-08-04T19:11:57.328830Z","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":"2026-08-04T19:11:57.333664Z","title":"Gradient descent with linearly correlated noise: Theory and applications to differential privacy","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":31,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.333664Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:906596a1e55163e573a9c5de2bdd739aba18759ca75f22cae72862d97be043b5","observation_id":"ff55522d-5d04-4080-8c5d-f4970401875c","resolution":{"observed_at":"2026-08-04T19:11:57.333664Z","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":"2026-08-04T19:11:57.338552Z","title":"Optimality of the johnson-lindenstrauss lemma","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":32,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.338552Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:f9be92576281f91da3533e5915bc3ff9ffa7741e6fa736ee13d57caa77ff1222","observation_id":"9cef15ce-754e-4b3b-8aac-d5179956fb78","resolution":{"observed_at":"2026-08-04T19:11:57.338552Z","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":"2026-08-04T19:11:57.343566Z","title":"Convergence and privacy of decentralized nonconvex optimization with gradient clipping and communication compression","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":33,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.343566Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:238666956c699522a933b4ee1aab8ea090f2b57dca818be5829e867ffd32937d","observation_id":"39e1b47f-788f-40ce-b2aa-1c18555a3cbc","resolution":{"observed_at":"2026-08-04T19:11:57.343566Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2210.13918","last_updated":"2022-10-26T12:55:53Z","snapshot_observed_at":"2026-08-16T16:21:20.418992Z","submitted_at":"2022-10-25T11:12:56Z","title":"Differentially Private Language Models for Secure Data Sharing","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2210.13918","snapshot_observed_at":"2026-08-04T19:11:57.348547Z","title":"Differentially private language models for secure data sharing","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":34,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.348547Z"},"links":{"cited_paper":"/paper/2210.13918","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:3c25d08c3972838414e0a039a76501b34aa3b4f6455f9b5ec57edd270401f5d6","observation_id":"9dcaf6ed-88bb-4bbf-b9b4-50fd0ed11784","resolution":{"observed_at":"2026-08-04T19:11:57.348547Z","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":"2026-08-04T19:11:57.353269Z","title":"Efficient deep learning: A survey on making deep learning models smaller, faster, and better","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":35,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.353269Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:e1a331cebabede987d91a19e41579ee0e1103f16420159ce08a889082d27662c","observation_id":"7e3a6ca2-d469-4231-9b4f-52fe7dfab084","resolution":{"observed_at":"2026-08-04T19:11:57.353269Z","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":"2026-08-04T19:11:57.359093Z","title":"Differentially private model compression","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":36,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.359093Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:ec5f24bcb7b1e57c2a5941111f637faf836703c9b0b93b0cb5a820d23e238114","observation_id":"6ec0a73f-2b7e-4bc5-99f9-f3d4ed511412","resolution":{"observed_at":"2026-08-04T19:11:57.359093Z","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":"2026-08-04T19:11:57.363826Z","title":"A survey of regularization strategies for deep models","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":37,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.363826Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:cde481f5d10b3bbb10769d3e354909866bf476d1a1876f704e80aae5a869eabc","observation_id":"03ab7929-86fa-4160-ae95-30e693beee1a","resolution":{"observed_at":"2026-08-04T19:11:57.363826Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1710.03163","last_updated":"2017-10-09T15:57:45Z","snapshot_observed_at":"2026-08-18T13:13:10.424401Z","submitted_at":"2017-10-09T15:57:45Z","title":"Random Projection and Its Applications","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1710.03163","snapshot_observed_at":"2026-08-04T19:11:57.368520Z","title":"Random projection and its applications","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":38,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.368520Z"},"links":{"cited_paper":"/paper/1710.03163","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:d4baf7a95c5c12b76ecbf7e8358220baee032632d265512611eb79367c9ae76b","observation_id":"8ffada95-9c57-4685-b4e8-5e9af912f81b","resolution":{"observed_at":"2026-08-04T19:11:57.368520Z","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":"2026-08-04T19:11:57.373752Z","title":"Explicit regularization in overparametrized models via noise injection","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":39,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.373752Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:822313de628cb06af1920ae709c73911add89b8076b017f3ac19c69adf99c6a4","observation_id":"101471ec-a8ae-4f6d-b07b-48003367bdce","resolution":{"observed_at":"2026-08-04T19:11:57.373752Z","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":"2026-08-04T19:11:57.378299Z","title":"How to dp-fy ml: A practical guide to machine learning with differential privacy","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":40,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.378299Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:5ed878693a6405654607cb755eac03ac41eea1748a3f4f17a3af6985714005fd","observation_id":"31578d37-8b9e-4744-8fd4-62fca8efceb6","resolution":{"observed_at":"2026-08-04T19:11:57.378299Z","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":"2026-08-04T19:11:57.382782Z","title":"Enabling fast differentially private sgd via just-in-time compilation and vectorization","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":41,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.382782Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:7fe00f5f6882ecc623251286c810636979316e9aa982027f2675d2f5e60fe637","observation_id":"863f23b1-569f-4843-9080-04a47b8d4185","resolution":{"observed_at":"2026-08-04T19:11:57.382782Z","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":"2026-08-04T19:11:57.388002Z","title":"Large language models in medicine","venue":null,"work_id":null,"year":1930},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":42,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.388002Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:f97bf8e37109a05227e16686da5e58d57f0fad7bcd77bcdb88f6032e4ac79dc0","observation_id":"322c5045-e065-4e69-b10e-24466f7e862c","resolution":{"observed_at":"2026-08-04T19:11:57.388002Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.06465","last_updated":"2024-10-25T19:05:43Z","snapshot_observed_at":"2026-08-16T14:19:52.109103Z","submitted_at":"2024-02-09T15:17:53Z","title":"On Differentially Private Subspace Estimation in a Distribution-Free Setting","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.06465","snapshot_observed_at":"2026-08-04T19:11:57.392746Z","title":"On differentially private subspace estimation without distributional assumptions","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":43,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.392746Z"},"links":{"cited_paper":"/paper/2402.06465","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:cbec6185e26dca196c812cbb2a892709aeaa1c40525dd63ee708e0fc7de72ca7","observation_id":"68a7a265-982f-45fa-8462-d0685ce54317","resolution":{"observed_at":"2026-08-04T19:11:57.392746Z","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":"2026-08-04T19:11:57.397802Z","title":"Resnets ensemble via the feynman-kac formalism to improve natural and robust accuracies","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":44,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.397802Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:77a577649f34a4be12730a05383c6ec75e55159becdb31ca7a548feaa134801c","observation_id":"e24b98c7-4070-4a70-b371-8371f09a5892","resolution":{"observed_at":"2026-08-04T19:11:57.397802Z","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":"2026-08-04T19:11:57.402638Z","title":"Private model compression via knowledge distillation","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":45,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.402638Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:abc202b13a142bc4975a4ac93bd36190349e2c535c9134ac1d8c7cdc13ca287e","observation_id":"d1f5be45-1cfc-4b81-b6bb-144bcac4ad8f","resolution":{"observed_at":"2026-08-04T19:11:57.402638Z","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":"2026-08-04T19:11:57.407544Z","title":"Protect privacy from gradient leakage attack in federated learning","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":46,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.407544Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:d7c6f96a4de7db393bf81f5257fa27671420fa53464de2a4c0c504b884ec669c","observation_id":"b253083f-d63f-4dd8-88a6-d3aceeef528b","resolution":{"observed_at":"2026-08-04T19:11:57.407544Z","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":"2026-08-04T19:11:57.412895Z","title":"Differentially private sgd with non-smooth losses","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":47,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.412895Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:2a3b07678747c76c43a941ea701243504db57e10ad34c1bf8e64b1b4c2bc44a2","observation_id":"a8936056-54f0-4dc5-9f5d-e68d06a4ae08","resolution":{"observed_at":"2026-08-04T19:11:57.412895Z","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":"2026-08-04T19:11:57.418341Z","title":"A theory to instruct differentially-private learning via clipping bias reduction","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":48,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.418341Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:bc9a1a90a60fba13547f48ceb73ffd31dac7aa42b14ed9c342261c64ea69f78b","observation_id":"1549615d-78dd-434e-b8d5-261e7a461de7","resolution":{"observed_at":"2026-08-04T19:11:57.418341Z","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":"2026-08-04T19:11:57.424086Z","title":"Robust regression and lasso","venue":null,"work_id":null,"year":2008},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":49,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.424086Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:2e1afab3aa800eb2c74748e3a2327ddb7e0393519b780cf27ed70d8b0b5de4a7","observation_id":"eada1a78-0b2f-44c5-99aa-c10802ad2a2d","resolution":{"observed_at":"2026-08-04T19:11:57.424086Z","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":"2026-08-16T17:54:08.474823Z","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":"2026-08-04T19:11:57.429770Z","title":"Opacus: User-friendly differential privacy library in pytorch","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":50,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.429770Z"},"links":{"cited_paper":"/paper/2109.12298","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:a9d95e44105ccaca693cffd53c802a3eeb8cc7ecbcffc8f63085e70cce77f3e7","observation_id":"d905cf6d-215d-4442-93a8-655192c64173","resolution":{"observed_at":"2026-08-04T19:11:57.429770Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2110.06500","last_updated":"2022-07-14T22:14:17Z","snapshot_observed_at":"2026-08-16T17:49:45.358995Z","submitted_at":"2021-10-13T05:15:00Z","title":"Differentially Private Fine-tuning of Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2110.06500","snapshot_observed_at":"2026-08-04T19:11:57.435191Z","title":"Differentially private fine-tuning of language models","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":51,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.435191Z"},"links":{"cited_paper":"/paper/2110.06500","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:15f2d5dc782103d1c6a1ea0d39a3cf60c0961d053c1eb62428f16cef3e0c1112","observation_id":"08b4b149-bc7a-4117-8013-6250178a2b9c","resolution":{"observed_at":"2026-08-04T19:11:57.435191Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2311.14632","last_updated":"2024-04-17T04:16:15Z","snapshot_observed_at":"2026-08-16T14:40:35.460676Z","submitted_at":"2023-11-24T17:56:44Z","title":"Differentially Private SGD Without Clipping Bias: An Error-Feedback Approach","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2311.14632","snapshot_observed_at":"2026-08-04T19:11:57.441113Z","title":"Differentially private sgd without clipping bias: An error-feedback approach","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":52,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.441113Z"},"links":{"cited_paper":"/paper/2311.14632","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:34e174bc7f4d695bbe7d60cc1760c57b8fb3e0910700dbd3c04bfe16d082a596","observation_id":"2940231d-a181-48c4-8d30-d758006b6f0c","resolution":{"observed_at":"2026-08-04T19:11:57.441113Z","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":"2026-08-18T08:58:00.011447Z","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":"2026-08-04T19:11:57.447593Z","title":"idlg: Improved deep leakage from gradients","venue":null,"work_id":null,"year":2001},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":53,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.447593Z"},"links":{"cited_paper":"/paper/2001.02610","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:8c2345584344ce3327f8bdb2ec5028b1387ebb74b1203b2e27b17dc419b6bd6b","observation_id":"b9b00dbc-8db5-447d-83b0-86992fbdb8e6","resolution":{"observed_at":"2026-08-04T19:11:57.447593Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2007.03813","last_updated":"2021-04-23T23:07:28Z","snapshot_observed_at":"2026-08-09T17:03:52.549604Z","submitted_at":"2020-07-07T22:31:01Z","title":"Bypassing the Ambient Dimension: Private SGD with Gradient Subspace Identification","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2007.03813","snapshot_observed_at":"2026-08-04T19:11:57.452693Z","title":"Bypassing the ambient dimension: Private sgd with gradient subspace identification","venue":null,"work_id":null,"year":2007},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":54,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.452693Z"},"links":{"cited_paper":"/paper/2007.03813","citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:7aad1596d217fb3201cfa086f529d0afa7e636f5e33ac9fbcff0d2ce73a13b7a","observation_id":"b7cea909-8d01-41d4-a7ee-57af05091880","resolution":{"observed_at":"2026-08-04T19:11:57.452693Z","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":"2026-08-04T19:11:57.458158Z","title":"write newline","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection","version":2},"reference_index":55,"source":"arxiv_source","source_observed_at":"2026-08-04T19:11:57.458158Z"},"links":{"citing_paper":"/paper/2509.09485"},"observation_digest":"sha256:653bbd553e29d2db73a0baa67d87c051554764f2dbb1d57cda5c65b63d9ac016","observation_id":"d2a90b3a-56cc-4082-8421-a772829efe53","resolution":{"observed_at":"2026-08-04T19:11:57.458158Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2509.09485","last_updated":"2025-09-12T01:27:15Z","latest_version":2,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-18T08:59:20.362811Z","submitted_at":"2025-09-11T14:17:04Z","title":"Balancing Utility and Privacy: Dynamically Private SGD with Random Projection"},"reference_resolution":{"displayed":55,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":55,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":55},"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":"2026-08-18T06:34:40.430872+00:00","source":"crossref"},{"observed_at":"2026-08-18T06:34:34.496301+00:00","source":"retraction_watch"}],"thesis":"As of 18 August 2026, this Paper Citation Record lists 55 of 55 outbound references and 0 inbound Pith citation observations for arXiv:2509.09485."}