{"as_of":"2026-08-17T19:15:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:d92960d8477052f25a24b2663a4edaa3ba9a927e5ab3319b48f0f531729e7ab2","coverage":[{"denominator":65,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":65,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-16T11:09:46.740169Z","state":"measured"},{"denominator":71,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":71,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-17T06:30:58.91139+00:00","state":"measured"},{"denominator":6,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":6,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-15T15:57:47.603991Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"arxiv_reference","source_observed_at":"2026-05-20T22:23:47.884213Z","state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2504.16438","snapshot_observed_at":"2026-08-08T19:09:03.980412Z","title":"Private federated learning using preference-optimized synthetic data","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2502.05505","last_updated":"2025-05-20T04:05:24Z","snapshot_observed_at":"2026-08-17T13:26:10.786750Z","submitted_at":"2025-02-08T09:50:30Z","title":"Differentially Private Synthetic Data via APIs 3: Using Simulators Instead of Foundation Model","version":3},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-08T19:09:03.980412Z"},"links":{"cited_paper":"/paper/2504.16438","citing_paper":"/paper/2502.05505"},"observation_digest":"sha256:13c6a45d6de04a97e7c528917a47a4cdc41e3eb78b940e27bc0570f87f1b8495","observation_id":"20f93e2d-e668-407f-8416-6b573769400c","resolution":{"observed_at":"2026-08-08T19:09:03.980412Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2504.16438","snapshot_observed_at":"2026-08-15T15:57:47.603991Z","title":"Private federated learning using preference-optimized synthetic data.arXiv preprint arXiv:2504.16438, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2509.10696","last_updated":"2025-09-12T21:18:13Z","snapshot_observed_at":"2026-08-16T14:50:00.678248Z","submitted_at":"2025-09-12T21:18:13Z","title":"Struct-Bench: A Benchmark for Differentially Private Structured Text Generation","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-15T15:57:47.603991Z"},"links":{"cited_paper":"/paper/2504.16438","citing_paper":"/paper/2509.10696"},"observation_digest":"sha256:b4816a41b5b62a83559ff0f3454844ff731d235d51d26f9c8b764ccd4db3e2a7","observation_id":"a911f33c-41a6-4f27-a5e9-a4c56fbc6c8c","resolution":{"observed_at":"2026-08-15T15:57:47.603991Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2504.16438","snapshot_observed_at":"2026-08-03T18:52:55.567821Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.03238","last_updated":"2026-07-09T21:40:53Z","snapshot_observed_at":"2026-08-14T04:28:16.108325Z","submitted_at":"2025-12-02T21:14:39Z","title":"How to DP-fy Your Data: A Practical Guide to Generating Synthetic Data With Differential Privacy","version":2},"reference_index":115,"source":"arxiv_source","source_observed_at":"2026-08-03T18:52:55.567821Z"},"links":{"cited_paper":"/paper/2504.16438","citing_paper":"/paper/2512.03238"},"observation_digest":"sha256:36cf7be97380542f63fe4dd09460fa0bc4d8e1fc5897ae9defb62013d5291a6f","observation_id":"2ad90ca6-42a2-48bd-bd3e-7bc100b6553d","resolution":{"observed_at":"2026-08-03T18:52:55.567821Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"cited_work":{"arxiv_id":"2504.16438","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2504.16438","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"arXiv preprint arXiv:2504.16438 , year=","venue":null,"work_id":"c7684dde-d5e1-4553-9ad8-72c73108210f","year":2025},"citing_paper":{"arxiv_id":"2604.18966","last_updated":"2026-05-17T12:00:22Z","snapshot_observed_at":"2026-08-16T15:11:03.089989Z","submitted_at":"2026-04-21T01:29:52Z","title":"Self-Improving Tabular Language Models via Iterative Reward-Guided Post-Training","version":1},"reference_index":6,"source":"arxiv_source","source_observed_at":"2026-05-10T03:04:54.146481Z"},"links":{"cited_paper":"/paper/2504.16438","citing_paper":"/paper/2604.18966"},"observation_digest":"sha256:23c27723463aced77900df84b541a7bdcaee1fbf51ce915910fe19e7904b98b0","observation_id":"5dfdeaec-9b27-49d9-88bb-2973dc6dbbe5","resolution":{"observed_at":"2026-05-11T12:46:04.078044Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"cited_work":{"arxiv_id":"2504.16438","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2504.16438","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"arXiv preprint arXiv:2504.16438 , year=","venue":null,"work_id":"c7684dde-d5e1-4553-9ad8-72c73108210f","year":2025},"citing_paper":{"arxiv_id":"2605.09855","last_updated":"2026-05-18T00:37:30Z","snapshot_observed_at":"2026-08-14T22:44:32.446622Z","submitted_at":"2026-05-11T01:17:58Z","title":"Concordia: Self-Improving Synthetic Tables for Federated LLMs","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-05-12T04:26:50.410397Z"},"links":{"cited_paper":"/paper/2504.16438","citing_paper":"/paper/2605.09855"},"observation_digest":"sha256:c5d836e5e6592891f9684afd6025af411ad65c5ac8fab04510118dfb325f1b56","observation_id":"44c3a53b-6e98-4135-9a3d-3560242e5b3f","resolution":{"observed_at":"2026-05-12T06:16:27.540255Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"cited_work":{"arxiv_id":"2504.16438","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2504.16438","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"arXiv preprint arXiv:2504.16438 , year=","venue":null,"work_id":"c7684dde-d5e1-4553-9ad8-72c73108210f","year":2025},"citing_paper":{"arxiv_id":"2605.09855","last_updated":"2026-05-18T00:37:30Z","snapshot_observed_at":"2026-08-14T22:44:32.446622Z","submitted_at":"2026-05-11T01:17:58Z","title":"Concordia: Self-Improving Synthetic Tables for Federated LLMs","version":2},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-05-20T22:21:03.637418Z"},"links":{"cited_paper":"/paper/2504.16438","citing_paper":"/paper/2605.09855"},"observation_digest":"sha256:cdcc3ddc5c1cd060c5be9af2bdea7492419b1ca8051d16ac33f7d961b9327cb8","observation_id":"cada88a5-020b-45e4-b14a-5d95000fea2b","resolution":{"observed_at":"2026-05-20T22:23:47.887440Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2504.16438/citation-record","integrity":"/paper/2504.16438/integrity","json":"/paper/2504.16438/citation-record.json","paper":"/paper/2504.16438"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T11:09:47.815474Z","title":"B., Mironov, I., Talwar, K., and Zhang, L","venue":null,"work_id":"2167f220-a4c6-4fde-a900-ee1dc4b47f68","year":2016},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":1,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.432239Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:3a33e38882909a7d3002b1eff407cb9d848be1e2518ba595135f1c5380fab095","observation_id":"4854c21f-613a-4e37-ab43-08fb490379d5","resolution":{"observed_at":"2026-08-16T11:09:47.820000Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2303.08774","last_updated":"2024-03-04T06:01:33Z","snapshot_observed_at":"2026-08-17T09:58:46.058102Z","submitted_at":"2023-03-15T17:15:04Z","title":"GPT-4 Technical Report","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.08774","snapshot_observed_at":"2026-08-16T11:09:46.438002Z","title":"L., Almeida, D., Altenschmidt, J., Altman, S., Anadkat, S., et al","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":2,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.438002Z"},"links":{"cited_paper":"/paper/2303.08774","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:eb2475127dcdea3845aa7fe68030be1fbd39ff3938ea89c2fc8e8871f0de9871","observation_id":"9c39decf-f7e2-477c-a709-31595e153ba6","resolution":{"observed_at":"2026-08-16T11:09:46.438002Z","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":"2026-08-16T11:09:46.443494Z","title":"Llama 3 model card","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":3,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.443494Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:0b7a843a50667661cc57ed8847e3c8a42fb6cf09cc0825444751343a5f4d4198","observation_id":"4fcb95a9-f933-4e98-9a2b-07c5263a20e6","resolution":{"observed_at":"2026-08-16T11:09:46.443494Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.10403","last_updated":"2023-09-13T20:35:45Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-05-17T17:46:53Z","title":"PaLM 2 Technical Report","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.10403","snapshot_observed_at":"2026-08-16T11:09:46.448072Z","title":"M., Firat, O., Johnson, M., Lepikhin, D., Passos, A., Shakeri, S., Taropa, E., Bailey, P., Chen, Z., et al","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":4,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.448072Z"},"links":{"cited_paper":"/paper/2305.10403","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:19e67aa54552b0d46c05e924afb0b11ffed8f7994a02f326075a5df25fa12cd4","observation_id":"c5bfd3a6-24a4-4cb9-a624-e45c51c038d9","resolution":{"observed_at":"2026-08-16T11:09:46.448072Z","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":"2026-08-16T11:09:47.790512Z","title":"J., Deems, S., Furlani, T","venue":null,"work_id":"ab68dbbd-3e24-4df3-9923-32257673c8b8","year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":5,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.454049Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:26eb76a6c82cccbc9805e59804be13a99fd08a11e39dac0bc08d84a5259b6af1","observation_id":"d7f3b586-372c-430b-aaab-59b768fd1306","resolution":{"observed_at":"2026-08-16T11:09:47.796020Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+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":"2026-08-16T11:09:47.776547Z","title":"Towards private synthetic text generation","venue":null,"work_id":"3155d1f3-e3bd-4977-a27e-a7b9bc0e7d21","year":2019},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":6,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.459697Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:c74fa8d6106e7b210ad319d273f950bf583cd549df41f251f5ab3dd9527fbf12","observation_id":"b83cf538-2252-418f-a334-f7d5b0ea6b59","resolution":{"observed_at":"2026-08-16T11:09:47.781245Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1611.04482","last_updated":"2016-11-14T17:14:55Z","snapshot_observed_at":"2026-08-16T16:24:08.139377Z","submitted_at":"2016-11-14T17:14:55Z","title":"Practical Secure Aggregation for Federated Learning on User-Held Data","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1611.04482","snapshot_observed_at":"2026-08-16T11:09:46.464937Z","title":"A., Ivanov, V., Kreuter, B., Marcedone, A., McMahan, H","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":7,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.464937Z"},"links":{"cited_paper":"/paper/1611.04482","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:be49b3a2a97149c925a793fe607cfc7fd9cf1445f066a74a1ee4cd48ae4aef43","observation_id":"e12ed300-5046-47fc-acc5-6c14b99e96c6","resolution":{"observed_at":"2026-08-16T11:09:46.464937Z","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":"2026-08-16T11:09:46.470371Z","title":"T., Buitrago, P., Hanna, E., Sanielevici, S., Scibek, R., and Nystrom, N","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":8,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.470371Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:c4e9398232873d14f27058c436a7fbf1260fcb1aedfe88e6620012d677726dc3","observation_id":"ee9cc9d8-d76a-490b-9f95-e560253506db","resolution":{"observed_at":"2026-08-16T11:09:46.470371Z","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":"biblio/1027487","doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-16T11:09:47.397904Z","title":null,"venue":null,"work_id":"ed48c98b-82bd-4646-ad5b-fbd980bf805b","year":2021},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":9,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.474854Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:95493d818a7666ae9f677ca46d029426c9b2153cdf9c37af15b917882f120264","observation_id":"1a8fab28-a908-471c-b1d5-7656eabf72a5","resolution":{"observed_at":"2026-08-16T11:09:47.405029Z","resolver_source":"raw_fallback","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2307.09619","last_updated":"2023-12-22T02:14:19Z","snapshot_observed_at":"2026-08-16T15:15:11.506034Z","submitted_at":"2023-07-18T20:27:45Z","title":"Towards Federated Foundation Models: Scalable Dataset Pipelines for Group-Structured Learning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2307.09619","snapshot_observed_at":"2026-08-16T11:09:46.479916Z","title":"Towards federated foundation models: Scalable dataset pipelines for group-structured learning","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":10,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.479916Z"},"links":{"cited_paper":"/paper/2307.09619","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:8ad5c1372adf67c4f15b0f07b2b9f4a926484fc27b325740a3b01c7a5686fd20","observation_id":"19fd856d-6d5f-4031-a2f4-7dc682801a46","resolution":{"observed_at":"2026-08-16T11:09:46.479916Z","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":"2026-08-16T11:09:47.762055Z","title":"B., Mitchell, N","venue":null,"work_id":"576ef94f-6407-42d0-8c4b-c3f5731dc280","year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":11,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.485296Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:60bbe7925373773142652e5e9e410aee8cf4f906eb3587b28fb1db0295ba4e58","observation_id":"5d228f39-3f46-4201-9d41-2a6e10abbd26","resolution":{"observed_at":"2026-08-16T11:09:47.766654Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2310.04627","last_updated":"2023-10-06T23:46:33Z","snapshot_observed_at":"2026-08-16T14:54:19.882709Z","submitted_at":"2023-10-06T23:46:33Z","title":"Profit: Benchmarking Personalization and Robustness Trade-off in Federated Prompt Tuning","version":1},"cited_work":{"arxiv_id":"2310.04627","doi":null,"metadata_source":"pith","pith_arxiv_id":"2310.04627","snapshot_observed_at":"2026-08-16T11:09:47.217596Z","title":"Profit: Benchmarking Personalization and Robustness Trade-off in Federated Prompt Tuning","venue":"cs.LG","work_id":"70ce2301-5596-402f-b8cc-f76bf40da41c","year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":12,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.490595Z"},"links":{"cited_paper":"/paper/2310.04627","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:d1a6700e0eb835f3a45bbf2c7a46a2aae54260df227f367e3fef88729431e7ad","observation_id":"a78f2629-68df-47b1-9ec6-db819a2facd0","resolution":{"observed_at":"2026-08-16T11:09:47.224741Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+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":"2026-08-16T11:09:46.495591Z","title":"Differential privacy","venue":null,"work_id":null,"year":2006},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":13,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.495591Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:8474fd39feec4175ba802258b2ce7c4e2c5ffcb4da709df11011b454acff0a30","observation_id":"9c02aa71-97f9-4bb3-be42-d7fce0be1800","resolution":{"observed_at":"2026-08-16T11:09:46.495591Z","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":"2026-08-16T11:09:46.500140Z","title":"and Roth, A","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":14,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.500140Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:4aa15025ce4cacc1872ee7eca84ea0a8b728c58e3dd6edcdae6c13688c55656d","observation_id":"6102c837-93c0-4b98-8b36-a56077a9ae28","resolution":{"observed_at":"2026-08-16T11:09:46.500140Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.21783","last_updated":"2024-11-23T23:27:33Z","snapshot_observed_at":"2026-08-13T17:20:44.002518Z","submitted_at":"2024-07-31T17:54:27Z","title":"The Llama 3 Herd of Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.21783","snapshot_observed_at":"2026-08-16T11:09:46.505937Z","title":"The llama 3 herd of models, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":15,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.505937Z"},"links":{"cited_paper":"/paper/2407.21783","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:cd34736467a33de4f69ce075fb6585b23842a588a2581b789d552f9307986c98","observation_id":"150c20f8-7ed2-41fa-b0e8-4440d70aafc4","resolution":{"observed_at":"2026-08-16T11:09:46.505937Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.12948","last_updated":"2026-01-04T03:57:36Z","snapshot_observed_at":"2026-08-15T12:33:55.451951Z","submitted_at":"2025-01-22T15:19:35Z","title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.12948","snapshot_observed_at":"2026-08-16T11:09:46.510467Z","title":"Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":16,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.510467Z"},"links":{"cited_paper":"/paper/2501.12948","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:50a57dbcd1435196a84563c37e3562b8021b236fea0e79c1c567cbc6c001190e","observation_id":"c74a4c8c-e12a-43c1-bc1f-3194b6162fec","resolution":{"observed_at":"2026-08-16T11:09:46.510467Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.04792","last_updated":"2024-02-29T20:59:17Z","snapshot_observed_at":"2026-08-16T14:20:38.839188Z","submitted_at":"2024-02-07T12:31:13Z","title":"Direct Language Model Alignment from Online AI Feedback","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.04792","snapshot_observed_at":"2026-08-16T11:09:46.514981Z","title":"Direct Language Model Alignment from Online AI Feedback","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":17,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.514981Z"},"links":{"cited_paper":"/paper/2402.04792","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:69776552b0310b7b4e783bc5788679a0652ee3662569f80b77b5c7a0ec28e65c","observation_id":"38fa54ce-2ca4-4ca5-95f2-0a989781cde4","resolution":{"observed_at":"2026-08-16T11:09:46.514981Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1811.03604","last_updated":"2019-02-28T21:07:51Z","snapshot_observed_at":"2026-08-15T21:55:07.573045Z","submitted_at":"2018-11-08T18:37:03Z","title":"Federated Learning for Mobile Keyboard Prediction","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1811.03604","snapshot_observed_at":"2026-08-16T11:09:46.520117Z","title":"Federated learning for mobile keyboard prediction, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":18,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.520117Z"},"links":{"cited_paper":"/paper/1811.03604","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:c9abe6e38d71cfc07a55c626c99a5871a66c08566277a425a9d3731bcbc8cc34","observation_id":"fadf57a7-df24-481c-b0c6-67f940dcf3fc","resolution":{"observed_at":"2026-08-16T11:09:46.520117Z","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":"2026-08-16T11:09:46.525597Z","title":"Gans trained by a two time-scale update rule converge to a local nash equilibrium","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":19,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.525597Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:a47660e89697c849351823aca64075557efa0c699e3f2e56dd6871a4640afc6c","observation_id":"86ae481a-cb76-451b-a013-7119ccac5a45","resolution":{"observed_at":"2026-08-16T11:09:46.525597Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.02958","last_updated":"2024-10-17T19:11:46Z","snapshot_observed_at":"2026-08-16T13:46:01.296675Z","submitted_at":"2024-06-05T05:27:02Z","title":"PrE-Text: Training Language Models on Private Federated Data in the Age of LLMs","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.02958","snapshot_observed_at":"2026-08-16T11:09:46.530911Z","title":"Pre-text: Training language models on private federated data in the age of llms, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":20,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.530911Z"},"links":{"cited_paper":"/paper/2406.02958","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:887fbd10d04d3f4611498a42276674486ec9bd097ee5ad9f5a73781601bb9e0b","observation_id":"b8e90d9e-3b2f-470e-b04d-985968c98a20","resolution":{"observed_at":"2026-08-16T11:09:46.530911Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2106.09685","last_updated":"2021-10-16T18:40:34Z","snapshot_observed_at":"2026-08-17T18:04:53.578114Z","submitted_at":"2021-06-17T17:37:18Z","title":"LoRA: Low-Rank Adaptation of Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2106.09685","snapshot_observed_at":"2026-08-16T11:09:46.535548Z","title":"J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., and Chen, W","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":21,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.535548Z"},"links":{"cited_paper":"/paper/2106.09685","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:abb1f6dbe8081de194d02a5691caa687e189cbf5c97f3b386f367f13a046bf82","observation_id":"c281c9dd-3444-4629-92be-1ca139803032","resolution":{"observed_at":"2026-08-16T11:09:46.535548Z","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":"2026-08-16T11:09:47.729206Z","title":"Practical and private (deep) learning without sampling or shuffling","venue":null,"work_id":"dfd05641-2a76-4b17-bcf8-e1cf6ef75547","year":2021},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":22,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.539867Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:0f297fa271cafb28ec8e9dbcbf28fbe7a6a3def9859b73d4672f71f8d73a21f0","observation_id":"004dd66d-7aa3-4f26-ac5d-56738d2ae7c4","resolution":{"observed_at":"2026-08-16T11:09:47.734353Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+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":"2026-08-16T11:09:47.714625Z","title":"B., Avent, B., Bellet, A., Bennis, M., Bhagoji, A","venue":null,"work_id":"2f28e2be-d435-4edc-a73a-411020c3ed7d","year":2021},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":23,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.543960Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:fd76b6f844f57677aaf0158291ffe815de77719a57557afa9b8699f9df759c44","observation_id":"ffe09adb-63a0-4763-bbd1-0e075f4cce5e","resolution":{"observed_at":"2026-08-16T11:09:47.719026Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2306.01684","last_updated":"2024-01-11T00:17:43Z","snapshot_observed_at":"2026-08-16T15:27:03.575086Z","submitted_at":"2023-06-02T16:59:36Z","title":"Harnessing large-language models to generate private synthetic text","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2306.01684","snapshot_observed_at":"2026-08-16T11:09:46.548155Z","title":"Harnessing large-language models to generate private synthetic text","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":24,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.548155Z"},"links":{"cited_paper":"/paper/2306.01684","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:d4aec2638c2a4edaee94141ab443e55199459246f0ec625942062022fc2e84d4","observation_id":"94155881-a85d-42a4-83ce-3b2a4585585a","resolution":{"observed_at":"2026-08-16T11:09:46.548155Z","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":"2026-08-16T11:09:47.699703Z","title":"On the privacy properties of gan-generated samples","venue":null,"work_id":"2bfc72c8-8f5b-46df-b732-507326196039","year":2021},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":25,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.552738Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:7828919bfc8b1c1b57841ebb25498a196c10a0811acdae5022483f8cb72b6a89","observation_id":"4ae090a7-86b6-433a-83fd-cd082fc68a6a","resolution":{"observed_at":"2026-08-16T11:09:47.705185Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.15560","last_updated":"2025-05-17T15:12:18Z","snapshot_observed_at":"2026-08-16T17:12:33.591572Z","submitted_at":"2023-05-24T23:47:26Z","title":"Differentially Private Synthetic Data via Foundation Model APIs 1: Images","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.15560","snapshot_observed_at":"2026-08-16T11:09:46.557484Z","title":"Differentially private synthetic data via foundation model apis 1: Images","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":26,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.557484Z"},"links":{"cited_paper":"/paper/2305.15560","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:29c025cdfac5372b4beb2f86666b7068140833bc8447b5d2e594237576ff4604","observation_id":"85f07afc-86db-48ae-ae3d-a97c01002213","resolution":{"observed_at":"2026-08-16T11:09:46.557484Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2502.05505","last_updated":"2025-05-20T04:05:24Z","snapshot_observed_at":"2026-08-17T13:26:10.786750Z","submitted_at":"2025-02-08T09:50:30Z","title":"Differentially Private Synthetic Data via APIs 3: Using Simulators Instead of Foundation Model","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2502.05505","snapshot_observed_at":"2026-08-16T11:09:46.562919Z","title":"Differentially private synthetic data via apis 3: Using simulators instead of foundation model, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":27,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.562919Z"},"links":{"cited_paper":"/paper/2502.05505","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:7e8084221aa819fdb3edf5b64ac855380f09bbb96d0ad303c42530f1e194a125","observation_id":"8165f8fa-5f6f-47a9-980f-8bc2840934e7","resolution":{"observed_at":"2026-08-16T11:09:46.562919Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2203.08242","last_updated":"2022-03-15T20:37:16Z","snapshot_observed_at":"2026-08-16T17:14:17.061433Z","submitted_at":"2022-03-15T20:37:16Z","title":"Data Contamination: From Memorization to Exploitation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2203.08242","snapshot_observed_at":"2026-08-16T11:09:46.567497Z","title":"and Schwartz, R","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":28,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.567497Z"},"links":{"cited_paper":"/paper/2203.08242","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:55282b57668e3050416f2b89ed05ad4fbcef5aa7a5c8223b99e0aba7bc7603f3","observation_id":"f8531fa0-b06d-4bda-abba-be6deed3678c","resolution":{"observed_at":"2026-08-16T11:09:46.567497Z","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":"2026-08-16T11:09:46.572583Z","title":"D., Chen, D., and Arora, S","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":29,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.572583Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:5ab9c9108f06a2375395f997cb5c2e4d4f343527cb9ba2308505bb076c14bdac","observation_id":"0ffdff74-7e9f-4df1-9de8-a8669845bcf6","resolution":{"observed_at":"2026-08-16T11:09:46.572583Z","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":"2026-08-16T11:09:46.577408Z","title":"Differentially private language models for secure data sharing","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":30,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.577408Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:5340bcef4c4fd44bffa976adef23d82cc91a736915a9e62baec92fe06ac6c65c","observation_id":"1103b5fc-2de2-4a25-9d68-14f25d0afebe","resolution":{"observed_at":"2026-08-16T11:09:46.577408Z","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":"2026-08-16T11:09:47.676492Z","title":null,"venue":null,"work_id":"412cce32-527e-4830-9c6f-ce9928445dc1","year":2017},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":31,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.581966Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:095f82b1e4c53d1724624d440cc06851e29ae4d3035afe64f807acdfde2a62c0","observation_id":"29bd471d-457d-4b4b-9682-cb74d5c2a235","resolution":{"observed_at":"2026-08-16T11:09:47.680886Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+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":"2026-08-16T11:09:47.660787Z","title":"Learning DP recurrent language models","venue":null,"work_id":"1ceb6f25-620b-47b5-abd5-d030dbddfaf5","year":2017},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":32,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.586518Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:fa3cefffaa0ff230a55f6f8b4ebe8d6129a05f0119cd845bf3be956bdd3a3071","observation_id":"a1eeb7a3-5ce6-4bd7-960b-ddb996e43662","resolution":{"observed_at":"2026-08-16T11:09:47.666196Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+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":"2026-08-16T11:09:47.645622Z","title":"A note on dpo with noisy preferences & relationship to ipo, November 25 2023","venue":null,"work_id":"a72ba9db-e7ac-467f-b840-bbf86eaa8c75","year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":33,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.590958Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:2741036fcbf7623c8a5c4a4bda15d6ae1f427509fc6ae145f5602760a99b0bd5","observation_id":"d7e1bc5a-21fc-4f91-831b-108c442b84ac","resolution":{"observed_at":"2026-08-16T11:09:47.651087Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.16938","last_updated":"2023-05-30T08:34:49Z","snapshot_observed_at":"2026-08-16T15:29:18.851413Z","submitted_at":"2023-05-26T13:55:17Z","title":"Few-shot Fine-tuning vs. In-context Learning: A Fair Comparison and Evaluation","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.16938","snapshot_observed_at":"2026-08-16T11:09:46.595498Z","title":"Few-shot fine-tuning vs","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":34,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.595498Z"},"links":{"cited_paper":"/paper/2305.16938","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:2e3c7b5f3a266e9b6b8c0f9f7e993285e92247a2146a7c275731d16672bdbbf3","observation_id":"730d0890-32bd-43ea-8caf-d688b7303af1","resolution":{"observed_at":"2026-08-16T11:09:46.595498Z","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":"2026-08-16T11:09:46.600875Z","title":"Federated learning with buffered asynchronous aggregation","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":35,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.600875Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:76d6592da4b8c57adfe7fd72299b25d46603575ef85c60af2f32d5be2412a2fa","observation_id":"41d7aaf3-68e4-4ed3-a7f8-4d49750b9907","resolution":{"observed_at":"2026-08-16T11:09:46.600875Z","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":"2026-08-16T11:09:46.605127Z","title":"Training language models to follow instructions with human feedback","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":36,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.605127Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:5a3388fe8cf70723cd312e21e2018a8fce62a971075bcc9149d1b8e0a6d33d3c","observation_id":"92ab5cdc-7cd2-4472-bb61-54d747483567","resolution":{"observed_at":"2026-08-16T11:09:46.605127Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2102.08503","last_updated":"2021-02-16T23:57:20Z","snapshot_observed_at":"2026-08-16T18:45:04.386558Z","submitted_at":"2021-02-16T23:57:20Z","title":"Federated Evaluation and Tuning for On-Device Personalization: System Design & Applications","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2102.08503","snapshot_observed_at":"2026-08-16T11:09:46.609318Z","title":null,"venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":37,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.609318Z"},"links":{"cited_paper":"/paper/2102.08503","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:949b1b216f3b0305e6070bc53e8f36240775f55dbfeeace9cf7ab8f2dc4e76bb","observation_id":"2fee7a39-6b05-49b9-a24c-7021464ecdbc","resolution":{"observed_at":"2026-08-16T11:09:46.609318Z","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":"2026-08-16T11:09:46.613800Z","title":"Language models are unsupervised multitask learners","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":38,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.613800Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:df69f970e346a45b6ac11c80de7074d7aa154c0ad02c0193b64592b10a30976b","observation_id":"1541da70-764a-40b0-aef5-98b39090efe1","resolution":{"observed_at":"2026-08-16T11:09:46.613800Z","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":"2026-08-16T11:09:46.618031Z","title":"D., Ermon, S., and Finn, C","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":39,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.618031Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:0e7f9bee19dbc1adf8d96143c97bc64ae6aafe6208f8deba91aa825dd36b6aac","observation_id":"ea8e97b4-1fea-411d-ab13-a9e216416993","resolution":{"observed_at":"2026-08-16T11:09:46.618031Z","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":"2026-08-16T11:09:46.622438Z","title":null,"venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":40,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.622438Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:fca3849f02a735e5ebb3cffa66bae2dcdb3a4963cd08e6bf95c6526a18861019","observation_id":"07041c73-a7ad-4a2d-b89e-4522507c77c6","resolution":{"observed_at":"2026-08-16T11:09:46.622438Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2003.00295","last_updated":"2021-09-08T23:37:17Z","snapshot_observed_at":"2026-08-13T06:06:10.844945Z","submitted_at":"2020-02-29T16:37:29Z","title":"Adaptive Federated Optimization","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2003.00295","snapshot_observed_at":"2026-08-16T11:09:46.626728Z","title":null,"venue":null,"work_id":null,"year":2003},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":41,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.626728Z"},"links":{"cited_paper":"/paper/2003.00295","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:e5b8f1efbbfb5891ef2e621b6fe53a758a5de3f5c68b4dd956fe12182895e1af","observation_id":"9748bbd1-1170-478b-b1f1-6297944d28c9","resolution":{"observed_at":"2026-08-16T11:09:46.626728Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1908.10084","last_updated":"2019-08-27T08:50:17Z","snapshot_observed_at":"2026-08-14T05:02:11.716316Z","submitted_at":"2019-08-27T08:50:17Z","title":"Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1908.10084","snapshot_observed_at":"2026-08-16T11:09:46.632355Z","title":"and Gurevych, I","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":42,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.632355Z"},"links":{"cited_paper":"/paper/1908.10084","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:752c19ad046e14182b2ecfd2d326a92f654ab6096fdc57ce2c8ff062113239b8","observation_id":"dc29e04f-0fe1-4765-a915-2f803192c1de","resolution":{"observed_at":"2026-08-16T11:09:46.632355Z","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":"2026-08-16T11:09:46.636974Z","title":"and Gurevych, I","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":43,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.636974Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:3e232198b8c31d4b1d149aa9bd8557abc052fcedd4bc69b14a0d01038c1bab14","observation_id":"f26e7fe0-c99c-4568-81a6-d050ab1b22f0","resolution":{"observed_at":"2026-08-16T11:09:46.636974Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.10628","last_updated":"2023-10-16T17:51:29Z","snapshot_observed_at":"2026-08-16T14:51:40.042976Z","submitted_at":"2023-10-16T17:51:29Z","title":"Data Contamination Through the Lens of Time","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.10628","snapshot_observed_at":"2026-08-16T11:09:46.641763Z","title":"Data contamination through the lens of time, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":44,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.641763Z"},"links":{"cited_paper":"/paper/2310.10628","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:ea425e7f634c6b11dd5de28e6c2a8b9ecaba8e277110ff9e25669a89364d3510","observation_id":"3135fdd4-e5f1-4ef9-af34-fd0ca9c374a1","resolution":{"observed_at":"2026-08-16T11:09:46.641763Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1910.01108","last_updated":"2020-03-01T02:57:50Z","snapshot_observed_at":"2026-08-07T19:07:36.327251Z","submitted_at":"2019-10-02T17:56:28Z","title":"DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1910.01108","snapshot_observed_at":"2026-08-16T11:09:46.646702Z","title":"Distilbert, a distilled version of bert: smaller, faster, cheaper and lighter","venue":null,"work_id":null,"year":1910},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":45,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.646702Z"},"links":{"cited_paper":"/paper/1910.01108","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:ef1b739f216c04a88733803c5ae2b307d79ee0700d11ddf7a8275d82d0503972","observation_id":"49e9d498-ff58-46a1-95dc-1cab0f9fa883","resolution":{"observed_at":"2026-08-16T11:09:46.646702Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2312.11805","last_updated":"2025-05-09T21:04:06Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-12-19T02:39:27Z","title":"Gemini: A Family of Highly Capable Multimodal Models","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2312.11805","snapshot_observed_at":"2026-08-16T11:09:46.652047Z","title":"M., Hauth, A., Millican, K., et al","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":46,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.652047Z"},"links":{"cited_paper":"/paper/2312.11805","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:993d1264b370d9da95e015664fac3daf72a1329e399afc4a2a4ae9df2027f093","observation_id":"7f4aa62f-1da2-4d75-9505-8c6c3d017048","resolution":{"observed_at":"2026-08-16T11:09:46.652047Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2302.13971","last_updated":"2023-02-27T17:11:15Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-02-27T17:11:15Z","title":"LLaMA: Open and Efficient Foundation Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2302.13971","snapshot_observed_at":"2026-08-16T11:09:46.656524Z","title":"Llama: Open and efficient foundation language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":47,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.656524Z"},"links":{"cited_paper":"/paper/2302.13971","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:e4aa33185bd13530e06902b4879f0fbcd22afaa282d31bc8005aaa34af134d30","observation_id":"a6770653-89c5-4ee0-b752-2c27e6d6481d","resolution":{"observed_at":"2026-08-16T11:09:46.656524Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1908.08962","last_updated":"2019-09-25T22:55:20Z","snapshot_observed_at":"2026-08-14T21:38:13.014467Z","submitted_at":"2019-08-23T18:02:05Z","title":"Well-Read Students Learn Better: On the Importance of Pre-training Compact Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1908.08962","snapshot_observed_at":"2026-08-16T11:09:46.660952Z","title":"Well-read students learn better: On the importance of pre-training compact models, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":48,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.660952Z"},"links":{"cited_paper":"/paper/1908.08962","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:39525129e779ba3104697ef2bbc0d5147bf9c20efbee89941541744656649d7a","observation_id":"f8e66126-29ac-4703-b372-7c4ddbefc9ff","resolution":{"observed_at":"2026-08-16T11:09:46.660952Z","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":"2026-08-16T11:09:46.666376Z","title":"Locally differentially private document generation using zero shot prompting","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":49,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.666376Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:8031001050d3ef8965642503d15e1d6b23bb00262fc247ba3c191f7a87e96ebf","observation_id":"f44ed986-2ed6-425d-8d1e-8370df9255ee","resolution":{"observed_at":"2026-08-16T11:09:46.666376Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2404.04360","last_updated":"2024-08-07T03:36:51Z","snapshot_observed_at":"2026-08-16T14:03:15.831800Z","submitted_at":"2024-04-05T19:14:14Z","title":"Prompt Public Large Language Models to Synthesize Data for Private On-device Applications","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2404.04360","snapshot_observed_at":"2026-08-16T11:09:46.670881Z","title":"Prompt public large language models to synthesize data for private on-device applications","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":50,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.670881Z"},"links":{"cited_paper":"/paper/2404.04360","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:41d18cc026c9111c5ec4b8200e4dbef900cbf46cc36feeb4549b30bd7b63f728","observation_id":"8d60ee5e-a656-49f5-aa6b-aa71ab2bd0c9","resolution":{"observed_at":"2026-08-16T11:09:46.670881Z","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":"2026-08-16T11:09:47.586829Z","title":"A., Nori, H., Jiang, H., Zhang, H., Lee, Y","venue":null,"work_id":"5af3e1e0-b2a8-491c-ab42-08f59efe0215","year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":51,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.675277Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:7d81f9f83227b98405cde67b4b2235923f9340d9978d6f43024f74b2caec5263","observation_id":"a20f4331-9ae5-4b07-99f1-aa00ef542ad2","resolution":{"observed_at":"2026-08-16T11:09:47.591340Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1802.06739","last_updated":"2018-02-19T18:05:33Z","snapshot_observed_at":"2026-08-14T19:44:27.939621Z","submitted_at":"2018-02-19T18:05:33Z","title":"Differentially Private Generative Adversarial Network","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1802.06739","snapshot_observed_at":"2026-08-16T11:09:46.679488Z","title":"Differentially private generative adversarial network","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":52,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.679488Z"},"links":{"cited_paper":"/paper/1802.06739","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:0d9056875fee6b2fb94e16f27d6ce87bc70c02c37101968105c4964dfa2fcb9f","observation_id":"a9dc0908-b145-489d-b83c-597b4b52da85","resolution":{"observed_at":"2026-08-16T11:09:46.679488Z","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":"2026-08-16T11:09:47.573621Z","title":null,"venue":null,"work_id":"968fe6d5-f2c4-44ba-ba16-0b5ee787a3bc","year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":53,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.684334Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:223300984d2e26203dd8d019c9dc0e3117e91a20516f5d0fad2a564abe58ea23","observation_id":"93d30527-d7f0-415d-a6cc-fce8e8b60087","resolution":{"observed_at":"2026-08-16T11:09:47.577812Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.18465","last_updated":"2023-07-17T08:28:34Z","snapshot_observed_at":"2026-08-16T15:28:46.157632Z","submitted_at":"2023-05-29T07:54:22Z","title":"Federated Learning of Gboard Language Models with Differential Privacy","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.18465","snapshot_observed_at":"2026-08-16T11:09:46.688638Z","title":"A., Kairouz, P., McMahan, H","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":54,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.688638Z"},"links":{"cited_paper":"/paper/2305.18465","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:37359aa77a7d6bcd307b148398784f4c065f446c0f816cbf23d5db6a25d567ed","observation_id":"944c672e-fc7c-412d-8267-94f177cfc3b0","resolution":{"observed_at":"2026-08-16T11:09:46.688638Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2311.04850","last_updated":"2023-11-11T05:11:18Z","snapshot_observed_at":"2026-08-16T14:44:53.286559Z","submitted_at":"2023-11-08T17:35:20Z","title":"Rethinking Benchmark and Contamination for Language Models with Rephrased Samples","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2311.04850","snapshot_observed_at":"2026-08-16T11:09:46.693633Z","title":"E., and Stoica, I","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":55,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.693633Z"},"links":{"cited_paper":"/paper/2311.04850","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:b46707b041d29ff5dae2fe9008833fbe501fd9979b7f7fd7de58e42bef97e1c2","observation_id":"281802d4-4a9d-4fd0-ae13-ad09cbb00999","resolution":{"observed_at":"2026-08-16T11:09:46.693633Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.04845","last_updated":"2024-06-07T11:19:30Z","snapshot_observed_at":"2026-08-16T13:45:11.269594Z","submitted_at":"2024-06-07T11:19:30Z","title":"FedLLM-Bench: Realistic Benchmarks for Federated Learning of Large Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.04845","snapshot_observed_at":"2026-08-16T11:09:46.698293Z","title":"Fedllm-bench: Realistic benchmarks for federated learning of large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":56,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.698293Z"},"links":{"cited_paper":"/paper/2406.04845","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:d531784a3102a95b5a38b9a99325cd63d4b8453de67a10abbd6679216a34821a","observation_id":"ed2c2a7e-a1db-4b16-b242-10f27cc7b105","resolution":{"observed_at":"2026-08-16T11:09:46.698293Z","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":"2026-08-16T11:09:47.559265Z","title":"Opacus: User-friendly differential privacy library in pytorch","venue":null,"work_id":"1e3f39e5-0cbc-43e1-964a-79b54793ab38","year":2021},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":57,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.702952Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:7bf9eae63dba1ee582645f68a371c263439c1a836a7afae57dcaedc6ab54daab","observation_id":"c747840c-a369-4e38-883e-fca0f5d7b1e9","resolution":{"observed_at":"2026-08-16T11:09:47.563621Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+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":"2026-08-16T11:09:47.544586Z","title":"Training private and efficient language models with synthetic data from LLM s","venue":null,"work_id":"57155bf7-0d6c-4e40-b8b5-7a535960cdaf","year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":58,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.707496Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:1bfe5013e4b3fdc384833ab4715aa1d72044b542418750a50ce0e46f7f594948","observation_id":"4a073d3e-462c-4a8f-bb00-3c4bf4b24f4e","resolution":{"observed_at":"2026-08-16T11:09:47.549628Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.13659","last_updated":"2024-07-02T11:57:07Z","snapshot_observed_at":"2026-08-16T14:16:40.146514Z","submitted_at":"2024-02-21T09:45:08Z","title":"Privacy-Preserving Instructions for Aligning Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.13659","snapshot_observed_at":"2026-08-16T11:09:46.711797Z","title":"Privacy-preserving instructions for aligning large language models, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":59,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.711797Z"},"links":{"cited_paper":"/paper/2402.13659","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:f27ae207d85b9dc8eb706df4cbfa96d8e1a07583224980d1c694968ea38a180d","observation_id":"c19db043-007f-4d3f-ace4-11b12a1cf907","resolution":{"observed_at":"2026-08-16T11:09:46.711797Z","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":"2026-08-16T11:09:46.716522Z","title":"Synthetic text generation with differential privacy: A simple and practical recipe","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":60,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.716522Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:6bd53d2e836ec5aed4eaf7c4478c84517e117ebb91d332fd351752ce0f9ad0bf","observation_id":"34d7da27-6800-46e1-a7f1-7bd87e997004","resolution":{"observed_at":"2026-08-16T11:09:46.716522Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2210.14348","last_updated":"2023-07-18T14:20:44Z","snapshot_observed_at":"2026-08-16T16:21:09.638453Z","submitted_at":"2022-10-25T21:21:17Z","title":"Synthetic Text Generation with Differential Privacy: A Simple and Practical Recipe","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2210.14348","snapshot_observed_at":"2026-08-16T11:09:46.721977Z","title":"A., Li, X., Kumar, G., McAnallen, J., Shajari, H., Sun, H., Levitan, D., and Sim, R","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":61,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.721977Z"},"links":{"cited_paper":"/paper/2210.14348","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:699a03ff2f4ba101217c3f2adb9ee5adc6542ad58de925b1cf4f4da73ac0b544","observation_id":"b6659a02-e3d3-4970-8868-804ed6f17b70","resolution":{"observed_at":"2026-08-16T11:09:46.721977Z","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":"2026-08-16T11:09:47.529222Z","title":"K., Oh, S., and He, N","venue":null,"work_id":"ef7ab6e1-a31d-4a5d-ad5a-c9f823a86b5b","year":2024},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":62,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.726585Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:3ea64cc4310f9216e3fefdcbec644c164c91c47e1237593f44cf912284b4c75b","observation_id":"a592f68b-fa82-4b63-864d-20df9e6a2ba1","resolution":{"observed_at":"2026-08-16T11:09:47.534780Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-17T06:30:58.91139+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2311.01964","last_updated":"2023-11-03T14:59:54Z","snapshot_observed_at":"2026-08-16T14:46:11.658740Z","submitted_at":"2023-11-03T14:59:54Z","title":"Don't Make Your LLM an Evaluation Benchmark Cheater","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2311.01964","snapshot_observed_at":"2026-08-16T11:09:46.730789Z","title":"X., Chen, X., Lin, Y., Wen, J.-R., and Han, J","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":63,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.730789Z"},"links":{"cited_paper":"/paper/2311.01964","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:bbcb03b3a4fc2f8392904caaf8e2a20d1c25f26fb02e09e7e6fc086f53487814","observation_id":"4123b1b4-6ccb-4f94-af8a-1543b0945bcf","resolution":{"observed_at":"2026-08-16T11:09:46.730789Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2502.00245","last_updated":"2025-02-01T00:54:25Z","snapshot_observed_at":"2026-08-09T23:06:19.225107Z","submitted_at":"2025-02-01T00:54:25Z","title":"Contrastive Private Data Synthesis via Weighted Multi-PLM Fusion","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2502.00245","snapshot_observed_at":"2026-08-16T11:09:46.735552Z","title":"Contrastive private data synthesis via weighted multi-plm fusion","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":64,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.735552Z"},"links":{"cited_paper":"/paper/2502.00245","citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:e7ca7a2f618337ef0882e7da08a2c02e6e92e5132a296a3a48ff91029b156b08","observation_id":"5a351d21-4159-4404-bd6f-3794aa5fd7a4","resolution":{"observed_at":"2026-08-16T11:09:46.735552Z","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":"2026-08-16T11:09:46.740169Z","title":"write newline","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data","version":2},"reference_index":65,"source":"arxiv_source","source_observed_at":"2026-08-16T11:09:46.740169Z"},"links":{"citing_paper":"/paper/2504.16438"},"observation_digest":"sha256:734de717da678bc6ecc30d712b7e1f22bad01ca395289e863b60ec4a31af01ff","observation_id":"18e905da-d2ed-490b-9a11-4b63c974404b","resolution":{"observed_at":"2026-08-16T11:09:46.740169Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2504.16438","last_updated":"2025-08-19T17:56:16Z","latest_version":2,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-16T14:49:12.099350Z","submitted_at":"2025-04-23T05:57:20Z","title":"POPri: Private Federated Learning using Preference-Optimized Synthetic Data"},"reference_resolution":{"displayed":65,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":50,"verified_exact":2,"verified_fuzzy":13},"total_outbound_references":65},"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":"2026-08-17T06:30:58.91139+00:00","source":"crossref"},{"observed_at":"2026-08-17T06:30:54.323127+00:00","source":"retraction_watch"}],"thesis":"As of 17 August 2026, this Paper Citation Record lists 65 of 65 outbound references and 6 inbound Pith citation observations for arXiv:2504.16438."}