{"as_of":"2026-08-19T20:17:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:c33092dddec0271e0bc19224bbaa3ff474e54fdd935f31fe4770f19cb3d1840c","coverage":[{"denominator":13,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":13,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-15T15:54:59.815140Z","state":"measured"},{"denominator":13,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":13,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-19T06:32:44.657259+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2509.21606/citation-record","integrity":"/paper/2509.21606/integrity","json":"/paper/2509.21606/citation-record.json","paper":"/paper/2509.21606"},"outbound":[{"citation":{"cited_paper":{"arxiv_id":"2306.15347","last_updated":"2023-06-27T10:00:06Z","snapshot_observed_at":"2026-08-19T16:54:05.576661Z","submitted_at":"2023-06-27T10:00:06Z","title":"FedET: A Communication-Efficient Federated Class-Incremental Learning Framework Based on Enhanced Transformer","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2306.15347","snapshot_observed_at":"2026-08-15T15:54:59.780237Z","title":"Adaptive plasticity improvement for continual learning","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.780237Z"},"links":{"cited_paper":"/paper/2306.15347","citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:16de55e60207e74485bb405017d2c7cf0f884cde66032be1ddb185332bd5d766","observation_id":"fab52fb0-309a-447e-80cc-3d5a776e4184","resolution":{"observed_at":"2026-08-15T15:54:59.780237Z","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":"2026-08-15T15:54:59.969891Z","title":"We set the task identity prediction threshold toϵl = 0.95,∀l , based on a hyperparameter search","venue":null,"work_id":"5708afbf-7fa6-4192-a65a-a8aeb590a442","year":2021},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.793119Z"},"links":{"citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:ea63f8a99b74ce9e1a581e595924b7381ee877fd49b62ad1f162707e6fd4990b","observation_id":"09321b81-d0da-4ba4-a843-caa405793122","resolution":{"observed_at":"2026-08-15T15:54:59.973325Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+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":"2026-08-15T15:54:59.958419Z","title":null,"venue":null,"work_id":"c595c8d5-7d38-4578-9030-5f5b57c58532","year":2019},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.796373Z"},"links":{"citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:765de28d92ae28a2b8a1779afefbe06a73d409558b6b06fc083de9cc97d72ca0","observation_id":"cd794cc5-b158-4b93-af37-de8ad36e4405","resolution":{"observed_at":"2026-08-15T15:54:59.962925Z","resolver_source":"raw_fallback","status":"malformed_identifier"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+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":"2026-08-15T15:54:59.937046Z","title":"A photo of a class","venue":null,"work_id":"48172896-eed0-46f3-9b38-859825db18e6","year":2021},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.803792Z"},"links":{"citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:68108ba308825c812f1a69e5e8e7d36e7a6f39b9ba00fdedcfe2ea7eddc77d5e","observation_id":"d6b411b7-fb03-46ec-8068-5f81aa5679a8","resolution":{"observed_at":"2026-08-15T15:54:59.940450Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+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":"2026-08-15T15:54:59.925715Z","title":"In our implementation, for each dataset we set the sampling dimension of the standard normal vector to five times the feature size","venue":null,"work_id":"36c94914-941f-4d5d-ab93-30a18042dd66","year":2021},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.807553Z"},"links":{"citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:41d21c6b399658dd8aa9222ee8d495bc67e2d6f6389079688a5a71aa7c10bc9c","observation_id":"45f2b17e-1dc1-47aa-a0db-515c10ca4cde","resolution":{"observed_at":"2026-08-15T15:54:59.929928Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+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":"2026-08-15T15:54:59.914862Z","title":"GLFC (Dong et al., 2022","venue":null,"work_id":"fe670cdb-20db-46e6-af5f-6f21993cf362","year":2022},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.811454Z"},"links":{"citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:74ff4e9ec43e73681709eab70df8a14d076ab72b1a2143bfa8146da274ca2a8e","observation_id":"10c1412f-63ec-4a74-b7d5-1a36066e3bd3","resolution":{"observed_at":"2026-08-15T15:54:59.918405Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+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":"2026-08-15T15:54:59.902637Z","title":"The follow-up studies (Liu et al., 2023; Dai et al., 2023; Li et al., 2024c;a) reduce the size of the replay cache but remain reliant upon old samples","venue":null,"work_id":"96d1c3f1-cc04-43d5-b206-1192ef8b3624","year":2023},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.815140Z"},"links":{"citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:291b99f4f8f9bd24ad7d219141077449089415238a3ac058700eef89b741558d","observation_id":"9c8b2f1e-7c66-4e31-b9ce-ccf74d960f97","resolution":{"observed_at":"2026-08-15T15:54:59.907660Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2412.13779","last_updated":"2025-05-15T03:29:15Z","snapshot_observed_at":"2026-08-14T07:04:51.642077Z","submitted_at":"2024-12-18T12:16:41Z","title":"Rehearsal-Free Continual Federated Learning with Synergistic Synaptic Intelligence","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.13779","snapshot_observed_at":"2026-08-15T15:54:59.776424Z","title":"Towards efficient replay in federated incremental learning","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":2019,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.776424Z"},"links":{"cited_paper":"/paper/2412.13779","citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:9868222ea1a505a0b455e7d9f813bac32ed40d568dd6286b15d766e0a18f22ee","observation_id":"13b87c8e-47df-4d9c-b9b7-90335eba15d7","resolution":{"observed_at":"2026-08-15T15:54:59.776424Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2405.00955","last_updated":"2024-05-02T02:33:15Z","snapshot_observed_at":"2026-08-19T16:53:52.040915Z","submitted_at":"2024-05-02T02:33:15Z","title":"Recovering Labels from Local Updates in Federated Learning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2405.00955","snapshot_observed_at":"2026-08-15T15:54:59.767855Z","title":"Recovering labels from local updates in federated learning.arXiv preprint arXiv:2405.00955,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":2020,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.767855Z"},"links":{"cited_paper":"/paper/2405.00955","citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:07bb44a870d5f715e2522637c82183d056437219bdbe3fc8b64728561a6ee7e9","observation_id":"f9f07b14-0e2b-41c1-8060-c3cb5c20d6ae","resolution":{"observed_at":"2026-08-15T15:54:59.767855Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2106.10270","last_updated":"2022-06-23T10:51:04Z","snapshot_observed_at":"2026-08-19T18:18:44.284335Z","submitted_at":"2021-06-18T17:58:20Z","title":"How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2106.10270","snapshot_observed_at":"2026-08-15T15:54:59.784900Z","title":"How to train your vit? data, augmentation, and regularization in vision transformers.arXiv preprint arXiv:2106.10270,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":2021,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.784900Z"},"links":{"cited_paper":"/paper/2106.10270","citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:9e802364cae56174f90c9523f467a945ff719745e06659a1094dd0edc23287b1","observation_id":"073f7bcc-7c39-4687-b9c6-af5f4e0df972","resolution":{"observed_at":"2026-08-15T15:54:59.784900Z","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":"2026-08-15T15:54:59.947428Z","title":"For CIFAR100, following the original paper we set the memory size to 2000; to satisfy memory constraints, for DomainNet and ImageNet-R the memory size is limited to","venue":null,"work_id":"70e83cc1-f20c-48a0-a4f3-bea2a0187fba","year":2000},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":2022,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.800274Z"},"links":{"citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:2e3441615c967e92361906d6309baa490a3238e70ebdd99ee9131b4bdb444c9d","observation_id":"f4af513e-d42f-48a1-a84f-59ef67dde270","resolution":{"observed_at":"2026-08-15T15:54:59.951290Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-19T06:32:44.657259+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2501.15718","last_updated":"2025-01-27T01:06:23Z","snapshot_observed_at":"2026-08-17T23:11:53.730745Z","submitted_at":"2025-01-27T01:06:23Z","title":"CENSOR: Defense Against Gradient Inversion via Orthogonal Subspace Bayesian Sampling","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.15718","snapshot_observed_at":"2026-08-15T15:54:59.789281Z","title":"Censor: Defense against gradient inversion via orthogonal subspace bayesian sampling.arXiv preprint arXiv:2501.15718,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":2023,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.789281Z"},"links":{"cited_paper":"/paper/2501.15718","citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:cad18bfee5cef6ffadf36cc4137976ac3d93cfeb6ebd2032b67509117c8954a5","observation_id":"8df439a0-cc18-4140-a48e-31b63eac32e7","resolution":{"observed_at":"2026-08-15T15:54:59.789281Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1907.02189","last_updated":"2020-06-25T06:45:52Z","snapshot_observed_at":"2026-08-14T16:08:56.121684Z","submitted_at":"2019-07-04T02:04:56Z","title":"On the Convergence of FedAvg on Non-IID Data","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1907.02189","snapshot_observed_at":"2026-08-15T15:54:59.772496Z","title":"On the convergence of fedavg on non-iid data.arXiv preprint arXiv:1907.02189,","venue":null,"work_id":null,"year":1907},"citing_paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection","version":3},"reference_index":2024,"source":"pdf_text","source_observed_at":"2026-08-15T15:54:59.772496Z"},"links":{"cited_paper":"/paper/1907.02189","citing_paper":"/paper/2509.21606"},"observation_digest":"sha256:92a3c5ec11289d849f5d0bc67d04f81c43ac15b7940400fa871ef35bb065948e","observation_id":"c2c01128-5d10-44b8-8be0-c754cb3386a2","resolution":{"observed_at":"2026-08-15T15:54:59.772496Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2509.21606","last_updated":"2026-06-21T17:57:37Z","latest_version":3,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-19T16:53:50.598665Z","submitted_at":"2025-09-25T21:20:00Z","title":"Task-Agnostic Federated Continual Learning via Replay-Free Gradient Projection"},"reference_resolution":{"displayed":13,"state_counts":{"malformed_identifier":1,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":6,"verified_exact":0,"verified_fuzzy":6},"total_outbound_references":13},"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":"2026-08-19T06:32:44.657259+00:00","source":"crossref"},{"observed_at":"2026-08-19T06:32:39.956319+00:00","source":"retraction_watch"}],"thesis":"As of 19 August 2026, this Paper Citation Record lists 13 of 13 outbound references and 0 inbound Pith citation observations for arXiv:2509.21606."}