{"as_of":"2026-08-22T06:44:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:a237a9bf929e3605edbaf58910c815fd19c684ceafc3b2a99e2fa2c1dc61ed55","coverage":[{"denominator":103,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":100,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-06T15:54:42.656667Z","state":"measured"},{"denominator":102,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":102,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-22T06:32:14.747728+00:00","state":"measured"},{"denominator":2,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":2,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-04T07:22:08.451533Z","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":"2026-05-19T15:17:39.326621Z","state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2507.14740","snapshot_observed_at":"2026-08-04T07:22:08.451533Z","title":"Better training data attribution via better inverse hessian-vector products","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2510.27313","last_updated":"2026-08-02T16:40:30Z","snapshot_observed_at":"2026-08-15T22:12:04.123851Z","submitted_at":"2025-10-31T09:39:12Z","title":"LLM generation novelty through the lens of semantic similarity","version":3},"reference_index":49,"source":"arxiv_source","source_observed_at":"2026-08-04T07:22:08.451533Z"},"links":{"cited_paper":"/paper/2507.14740","citing_paper":"/paper/2510.27313"},"observation_digest":"sha256:a1bd344dab36cb56c102015c95f1e6dff2330147b92c3860c6f961a84443563c","observation_id":"374aeb14-73c7-4a0f-82e9-2db62305992b","resolution":{"observed_at":"2026-08-04T07:22:08.451533Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"cited_work":{"arxiv_id":"2507.14740","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2507.14740","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"A., and Grosse, R","venue":null,"work_id":"95c8dd23-233c-41c6-ba18-58fc9061418c","year":null},"citing_paper":{"arxiv_id":"2605.15520","last_updated":"2026-05-15T01:34:55Z","snapshot_observed_at":"2026-08-15T03:49:13.229897Z","submitted_at":"2026-05-15T01:34:55Z","title":"On the Fragility of Data Attribution When Learning Is Distributed","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-05-19T15:16:32.125215Z"},"links":{"cited_paper":"/paper/2507.14740","citing_paper":"/paper/2605.15520"},"observation_digest":"sha256:690633853577f8d04d5044a003f71bbc52df3638de96aec60d6b636122ca6429","observation_id":"cec07009-02db-4c5a-9004-ea2736184a8a","resolution":{"observed_at":"2026-05-19T15:17:39.328500Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2507.14740/citation-record","integrity":"/paper/2507.14740/integrity","json":"/paper/2507.14740/citation-record.json","paper":"/paper/2507.14740"},"outbound":[{"citation":{"cited_paper":{"arxiv_id":"2405.13954","last_updated":"2024-05-22T19:39:05Z","snapshot_observed_at":"2026-08-20T18:10:29.535542Z","submitted_at":"2024-05-22T19:39:05Z","title":"What is Your Data Worth to GPT? LLM-Scale Data Valuation with Influence Functions","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2405.13954","snapshot_observed_at":"2026-08-06T15:54:42.360721Z","title":"What is your data worth to gpt? llm-scale data valuation with influence functions.arXiv preprint arXiv:2405.13954, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.360721Z"},"links":{"cited_paper":"/paper/2405.13954","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:585551acc4d05267e8de8be33303b969784259d4ed4699e54ed062eac7e69bb4","observation_id":"4d6c20fe-90c0-45fb-9f84-3884d85ce1ea","resolution":{"observed_at":"2026-08-06T15:54:42.360721Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.364870Z","title":"Improving subgroup robustness via data selection.Advances in Neural Information Processing Systems, 37:94490–94511, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.364870Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:4553b97a8bdc3ce88a07e603948a70ddcfd35a8c19e88ce5a294674139804621","observation_id":"7fff5dfa-c60a-483f-b3f8-091b63c9a5f4","resolution":{"observed_at":"2026-08-06T15:54:42.364870Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.368007Z","title":"Interactive label cleaning with example-based explanations.Advances in Neural Information Processing Systems, 34:12966–12977, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.368007Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:78edb02f9a433c4a1c57acd470afe5a1c652860a1f9b411c2f665c5800c2bc9d","observation_id":"da99d8d8-7d65-4cc0-86b2-1fe335194367","resolution":{"observed_at":"2026-08-06T15:54:42.368007Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.371040Z","title":"Data cleansing for models trained with sgd.Advances in Neural Information Processing Systems, 32, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.371040Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:5fbed5788df7c62088a7f81e7c56e6c350bd8fc26b50b04f57e929b51e36d184","observation_id":"71cf3544-6636-4d42-812c-f44d1edac0bf","resolution":{"observed_at":"2026-08-06T15:54:42.371040Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.374016Z","title":"On the accuracy of influence functions for measuring group effects.Advances in neural information processing systems, 32, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.374016Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:7bea892407be942d0b95e4613ba6902fac8038b88a1820381f2df9d6f99d26b9","observation_id":"75abe443-9a4b-44f7-97f3-e1cec7fad187","resolution":{"observed_at":"2026-08-06T15:54:42.374016Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.377120Z","title":"Representer point selection for explaining deep neural networks.Advances in neural information processing systems, 31, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.377120Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:68a607f4895919c80be1b9bb63ecea8a00848331dcfcef74241df5f6a2cc6892","observation_id":"c4f04e8c-d519-4011-bfe2-d2ed888e7ca6","resolution":{"observed_at":"2026-08-06T15:54:42.377120Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2308.03296","last_updated":"2023-08-07T04:47:42Z","snapshot_observed_at":"2026-08-19T20:23:36.050723Z","submitted_at":"2023-08-07T04:47:42Z","title":"Studying Large Language Model Generalization with Influence Functions","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2308.03296","snapshot_observed_at":"2026-08-06T15:54:42.380502Z","title":null,"venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.380502Z"},"links":{"cited_paper":"/paper/2308.03296","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:cb14e132a5c0dcce0537e032814d571c0e325a527b5e55a7b627131627bf32fc","observation_id":"4d447323-e630-4922-bd9e-c3b3b2780e30","resolution":{"observed_at":"2026-08-06T15:54:42.380502Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2205.11482","last_updated":"2022-10-25T12:45:15Z","snapshot_observed_at":"2026-08-21T04:23:46.745239Z","submitted_at":"2022-05-23T17:34:16Z","title":"Towards Tracing Factual Knowledge in Language Models Back to the Training Data","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2205.11482","snapshot_observed_at":"2026-08-06T15:54:42.384563Z","title":"Tracing knowledge in language models back to the training data.arXiv preprint arXiv:2205.11482, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.384563Z"},"links":{"cited_paper":"/paper/2205.11482","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:43cf589784806ff4f70be88268c80ba889e8bbfcd1c8e72c35cb87b65282c6a6","observation_id":"0881a6a2-38ed-41fb-bc7f-85e9c7b23cc1","resolution":{"observed_at":"2026-08-06T15:54:42.384563Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.387510Z","title":"Error discovery by clustering influence embeddings.Advances in Neural Information Processing Systems, 36:41765–41777, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.387510Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:db7824497aff5ea8d532b593e51a1f0da3ed8f07db1929b743093b85288da42f","observation_id":"69a9a371-8b61-48b7-add5-0612ba8a4dfc","resolution":{"observed_at":"2026-08-06T15:54:42.387510Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2202.00622","last_updated":"2022-02-01T18:15:24Z","snapshot_observed_at":"2026-08-20T14:38:32.002379Z","submitted_at":"2022-02-01T18:15:24Z","title":"Datamodels: Predicting Predictions from Training Data","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2202.00622","snapshot_observed_at":"2026-08-06T15:54:42.390773Z","title":"Datamodels: Predicting predictions from training data.arXiv preprint arXiv:2202.00622, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.390773Z"},"links":{"cited_paper":"/paper/2202.00622","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:e3fee17d16e3cc755b6431eb37f81828bd4c58ec69ca5495d2fbce3d88fbbed8","observation_id":"81abf800-ee1b-4e0c-8b17-9521223c62bf","resolution":{"observed_at":"2026-08-06T15:54:42.390773Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2410.13850","last_updated":"2025-05-24T14:59:10Z","snapshot_observed_at":"2026-08-19T05:50:42.676235Z","submitted_at":"2024-10-17T17:59:02Z","title":"Influence Functions for Scalable Data Attribution in Diffusion Models","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2410.13850","snapshot_observed_at":"2026-08-06T15:54:42.393828Z","title":"Influence functions for scalable data attribution in diffusion models.arXiv preprint arXiv:2410.13850, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.393828Z"},"links":{"cited_paper":"/paper/2410.13850","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:cb93f20aa3497b44b919cb51441e24e2398b0de3b2f9c7be2b5e822b87212242","observation_id":"de0e8e57-256d-49b2-8559-7bdf3abaa7c1","resolution":{"observed_at":"2026-08-06T15:54:42.393828Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.397887Z","title":"Understanding the origins of bias in word embeddings","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.397887Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:946393107fc1ee239be67394d972d967a3fff36397fdbf97f13f57734cbb5b3c","observation_id":"9f9ba4a5-4cfa-41ca-9218-16343955cb86","resolution":{"observed_at":"2026-08-06T15:54:42.397887Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.400796Z","title":"Fairif: Boosting fairness in deep learning via influence functions with validation set sensitive attributes","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.400796Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:fd7b694967ac7fdc0112a28cbd31ed95876b386463ecd77a404698cd83caa06b","observation_id":"295ab03a-b955-48f8-af3f-3a55aa6cebc7","resolution":{"observed_at":"2026-08-06T15:54:42.400796Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.403593Z","title":"On memorization in probabilistic deep generative models.Advances in Neural Information Processing Systems, 34:27916–27928, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.403593Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:9f87eb8704ed44117aecb66000201e02625c8e0f434618f08a7f7d27bd8185de","observation_id":"f39c9a40-f1d0-479d-8880-10c79e46704f","resolution":{"observed_at":"2026-08-06T15:54:42.403593Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2504.01032","last_updated":"2025-03-29T18:08:04Z","snapshot_observed_at":"2026-08-16T12:45:42.769126Z","submitted_at":"2025-03-29T18:08:04Z","title":"Who Owns the Output? Bridging Law and Technology in LLMs Attribution","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2504.01032","snapshot_observed_at":"2026-08-06T15:54:42.406412Z","title":"Who owns the output? bridging law and technology in llms attribution.arXiv preprint arXiv:2504.01032, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.406412Z"},"links":{"cited_paper":"/paper/2504.01032","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:6b26916b72da24a1d1300783ad429ff05242d2761c969b0917fec9b7aaf66cfc","observation_id":"fabade7f-c400-43c9-b870-d243d4127910","resolution":{"observed_at":"2026-08-06T15:54:42.406412Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.409571Z","title":"The influence curve and its role in robust estimation.Journal of the american statistical association, 69(346):383–393, 1974","venue":null,"work_id":null,"year":1974},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.409571Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:36756078c3be32c9737ff39a7a0b06a0313e2dbe65c87b646d0ab39fea16367d","observation_id":"44dedcd0-87cf-46b7-a16c-ad91e19f6f1a","resolution":{"observed_at":"2026-08-06T15:54:42.409571Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.412247Z","title":"Understanding black-box predictions via influence functions","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.412247Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:a4948277ec17158d835581c308645aaec69586f93b849f9eab2a09b4cdbd114c","observation_id":"44ea3784-e669-4319-b208-d2637c8a8016","resolution":{"observed_at":"2026-08-06T15:54:42.412247Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.415257Z","title":"Hydra: Hypergradi- ent data relevance analysis for interpreting deep neural networks","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.415257Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:fac6706e274bd1d2babd1572a9d17099a32608d395f8e8db5a0a0a17fdb5876f","observation_id":"e295c1d7-2b0b-4d02-a6fe-4dd0e375bae1","resolution":{"observed_at":"2026-08-06T15:54:42.415257Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2405.12186","last_updated":"2024-05-21T04:26:45Z","snapshot_observed_at":"2026-08-20T12:15:48.674343Z","submitted_at":"2024-05-20T17:17:44Z","title":"Training Data Attribution via Approximate Unrolled Differentiation","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2405.12186","snapshot_observed_at":"2026-08-06T15:54:42.418110Z","title":"Training data attribution via approximate unrolled differentation.arXiv preprint arXiv:2405.12186, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.418110Z"},"links":{"cited_paper":"/paper/2405.12186","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:cb603b90abd02ecdfbc838f9932e6014415472218dc0c2d1ed4b3c297744b4e4","observation_id":"c1800acf-baf2-4dcb-9391-ed2922cc871b","resolution":{"observed_at":"2026-08-06T15:54:42.418110Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.09538","last_updated":"2024-12-12T18:28:55Z","snapshot_observed_at":"2026-08-18T08:34:33.072285Z","submitted_at":"2024-12-12T18:28:55Z","title":"Capturing the Temporal Dependence of Training Data Influence","version":1},"cited_work":{"arxiv_id":"2412.09538","doi":null,"metadata_source":"pith","pith_arxiv_id":"2412.09538","snapshot_observed_at":"2026-08-06T15:54:42.929433Z","title":"Capturing the Temporal Dependence of Training Data Influence","venue":"cs.LG","work_id":"5190d65f-3a46-4ae4-bb86-8018de604ca7","year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.421086Z"},"links":{"cited_paper":"/paper/2412.09538","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:0866e505d405a56b05126a4d34fd1d3b4cb260c7acbe003fc109a2b832b0bff0","observation_id":"1ae4da80-2c05-45fd-a668-02abaf49840a","resolution":{"observed_at":"2026-08-06T15:54:42.932940Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2504.16430","last_updated":"2025-04-23T05:32:37Z","snapshot_observed_at":"2026-08-21T15:23:51.438903Z","submitted_at":"2025-04-23T05:32:37Z","title":"MAGIC: Near-Optimal Data Attribution for Deep Learning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2504.16430","snapshot_observed_at":"2026-08-06T15:54:42.424199Z","title":"Magic: Near-optimal data attribution for deep learning","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.424199Z"},"links":{"cited_paper":"/paper/2504.16430","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:8066888558ae4689053c59cbc5e072ff6457ed2f08e4d81b773d0c37c26b2ca1","observation_id":"4c7db193-a0bf-4e0e-85a7-89e3682e91f4","resolution":{"observed_at":"2026-08-06T15:54:42.424199Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2303.14186","last_updated":"2023-04-03T17:37:50Z","snapshot_observed_at":"2026-08-18T07:05:12.412330Z","submitted_at":"2023-03-24T17:56:22Z","title":"TRAK: Attributing Model Behavior at Scale","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.14186","snapshot_observed_at":"2026-08-06T15:54:42.427265Z","title":"Trak: Attributing model behavior at scale.arXiv preprint arXiv:2303.14186, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.427265Z"},"links":{"cited_paper":"/paper/2303.14186","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:d9b82877f1212028964489e44a3c3ea13656fdb118718fb31f0da97ee607f237","observation_id":"0386e0a6-6132-4f45-b9ac-d3e55057f9e7","resolution":{"observed_at":"2026-08-06T15:54:42.427265Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.430463Z","title":"Second-order stochastic optimization for machine learning in linear time.Journal of Machine Learning Research, 18(116):1–40, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.430463Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:079a3466238c77cf2854af30a15e66545cab50240a6fc6970da39174245df963","observation_id":"ce856ab6-9b65-4ddb-a04b-0bb8533528c5","resolution":{"observed_at":"2026-08-06T15:54:42.430463Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.433537Z","title":"Cambridge university press, 2012","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.433537Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:234768e8ec5d82e82e28c8c00547e244d633d739e81f9d4e8362ded805b44232","observation_id":"75d95e62-7f7d-426b-92b0-fada077e5ceb","resolution":{"observed_at":"2026-08-06T15:54:42.433537Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.436396Z","title":"Optimizing millions of hyperparameters by implicit differentiation","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.436396Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:9d6fc7f51f30aabc174e0f9dc2a133bd84fd262b9190db106825072ef6480d8d","observation_id":"c3f7b056-fa0f-4d02-b03d-94a99f3b7aaa","resolution":{"observed_at":"2026-08-06T15:54:42.436396Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.439403Z","title":"Optimizing neural networks with kronecker-factored approximate curvature","venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.439403Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:5fde3caadd4de30796aed05e3aa677368859c1d980391b3f80ca13b3481a8c3f","observation_id":"aefacf57-6c14-4732-a377-dbef43cb3c3a","resolution":{"observed_at":"2026-08-06T15:54:42.439403Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.442262Z","title":"Fast approximate natural gradient descent in a kronecker factored eigenbasis.Advances in Neural Information Processing Systems, 31, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.442262Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:1fdffd0258c256f90058ee7503ab7243d093d3ed17abeb1f104f934a59724180","observation_id":"c1916587-314d-4066-bc8b-347fa7f14861","resolution":{"observed_at":"2026-08-06T15:54:42.442262Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.445138Z","title":"Wright.Numerical optimization","venue":null,"work_id":null,"year":2006},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.445138Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:184c81e53bf69a27e1ee0644279dd17062f624ca3a0f6c697d6e28b28d23bc8e","observation_id":"f95622b3-af45-4ead-9730-1995dbfa69bf","resolution":{"observed_at":"2026-08-06T15:54:42.445138Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.448014Z","title":"On implicit bias in overparameterized bilevel optimization","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.448014Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:a8b6f7ab1cd155d93b77b6e6c44fce06f09fa1b8246cfe4ab5e7ce9efe98f72d","observation_id":"303534d3-1afe-4093-b52d-0ab895000300","resolution":{"observed_at":"2026-08-06T15:54:42.448014Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1611.07476","last_updated":"2017-10-05T13:28:50Z","snapshot_observed_at":"2026-08-20T15:31:18.469129Z","submitted_at":"2016-11-22T19:24:49Z","title":"Eigenvalues of the Hessian in Deep Learning: Singularity and Beyond","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1611.07476","snapshot_observed_at":"2026-08-06T15:54:42.450866Z","title":"Eigenvalues of the hessian in deep learning: Singularity and beyond.arXiv preprint arXiv:1611.07476, 2016","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.450866Z"},"links":{"cited_paper":"/paper/1611.07476","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:e6aab606602d8d39f0f804bb975ce0dff89377ba95721397daf6eb22bb66463b","observation_id":"8163ee90-3146-4e2d-bf0f-beb3df5b5b1a","resolution":{"observed_at":"2026-08-06T15:54:42.450866Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.453926Z","title":"An investigation into neural net optimization via hessian eigenvalue density","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.453926Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:9376db6f0adde002b2b0f07c3d853e258719c983a74728db6b34115c340d0b5f","observation_id":"f4ab4e84-dda6-4263-ba54-6e998ae74b97","resolution":{"observed_at":"2026-08-06T15:54:42.453926Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2409.17357","last_updated":"2024-09-25T21:05:35Z","snapshot_observed_at":"2026-08-20T00:17:41.097827Z","submitted_at":"2024-09-25T21:05:35Z","title":"Revisiting inverse Hessian vector products for calculating influence functions","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.17357","snapshot_observed_at":"2026-08-06T15:54:42.456707Z","title":"Revisiting inverse hessian vector products for calculating influence functions.arXiv preprint arXiv:2409.17357, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.456707Z"},"links":{"cited_paper":"/paper/2409.17357","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:fb81ff997b601e5fe2be454a6b3975ef118d032c31e1b03f328996efac8e3ef0","observation_id":"329ee866-d212-4a55-8547-b5d32bd1c27d","resolution":{"observed_at":"2026-08-06T15:54:42.456707Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.459900Z","title":"Scaling up influence functions","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.459900Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:feef7e9b36b58723ccb947080bfbe556f0493665596907807c614e93c018b562","observation_id":"3f31072a-bae9-4352-a312-49db06dd329f","resolution":{"observed_at":"2026-08-06T15:54:42.459900Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2006.14651","last_updated":"2021-02-10T23:45:14Z","snapshot_observed_at":"2026-08-21T05:29:09.642579Z","submitted_at":"2020-06-25T18:25:59Z","title":"Influence Functions in Deep Learning Are Fragile","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2006.14651","snapshot_observed_at":"2026-08-06T15:54:42.462806Z","title":"Influence functions in deep learning are fragile.arXiv preprint arXiv:2006.14651, 2020","venue":null,"work_id":null,"year":2006},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.462806Z"},"links":{"cited_paper":"/paper/2006.14651","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:d6e8d235dc52f068d1f2057b4b880945c57b3edd7155e4eead0f0b7a213d5455","observation_id":"155ca38e-bf19-4a8b-8f59-fb0c86678562","resolution":{"observed_at":"2026-08-06T15:54:42.462806Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.466017Z","title":"Theoretical and practical perspectives on what influence functions do.Advances in Neural Information Processing Systems, 36, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.466017Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:ce446dcd8979ad6b30e24176bb3dca68e296649570dd23f56d9fe59cad4e9ec7","observation_id":"a05f107e-63c8-49ff-94cf-1b571e873169","resolution":{"observed_at":"2026-08-06T15:54:42.466017Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.468775Z","title":"If influence functions are the answer, then what is the question?Advances in Neural Information Processing Systems, 35:17953–17967, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.468775Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:286dc03899dbf9a1f2131cf0266e2d3b119e9f558747c2c5bda7e4cae70c91a6","observation_id":"164bf317-25e9-458a-91ac-b88d57da7a9c","resolution":{"observed_at":"2026-08-06T15:54:42.468775Z","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":"2026-08-06T15:54:43.463079Z","title":"The proof and measurement of association between two things.The American Journal of Psychology, 15(1):72–101, 1904","venue":null,"work_id":"c29f45c0-5250-4133-99c2-e02b4e4c330c","year":1904},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.471706Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:01b8208f0050b6daee9cfbc2d115f5b8318a6bbf6b4e5bae34604a6cb0e687b8","observation_id":"7dc9b1d0-a837-43af-b101-2a644863ef12","resolution":{"observed_at":"2026-08-06T15:54:43.466272Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.454127Z","title":"Spectral algorithms for supervised learning.Neural Computation, 20(7):1873–1897, 2008","venue":null,"work_id":"121612c7-690a-41e8-b87f-555b57ec3887","year":2008},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.474512Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:49433626ae8bfb486c64e7a48b0d9f2c642fdf052bb5e5b5764462096bc1acc4","observation_id":"cde68334-25b5-4e7a-83cd-28b776595bfb","resolution":{"observed_at":"2026-08-06T15:54:43.457346Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.477413Z","title":"Understanding the difficulty of training deep feedforward neural networks","venue":null,"work_id":null,"year":2010},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.477413Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:166aebd46ad988be4e012f8936fbfba8a962e34ecf9e9be86d1df7caccbb1048","observation_id":"6b2b7934-af4b-40ba-bf83-177c57c6f527","resolution":{"observed_at":"2026-08-06T15:54:42.477413Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.480277Z","title":"Dropout: a simple way to prevent neural networks from overfitting.The journal of machine learning research, 15(1):1929–1958, 2014","venue":null,"work_id":null,"year":1929},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.480277Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:2edaae02194aa00cbdaf38c88968392627bdbf89e33993442f33f3e6e16bda6d","observation_id":"6f1078a3-c195-4e4c-b793-4255944bbf12","resolution":{"observed_at":"2026-08-06T15:54:42.480277Z","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":"2026-08-06T15:54:43.434378Z","title":"Efficient mini-batch training for stochastic optimization","venue":null,"work_id":"441f5a5f-975e-4477-9d4d-68d4707be73e","year":2014},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.483266Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:39921d2a17f57a7b16b048d448040a27ad02582dc92af74167a96936c3d038f1","observation_id":"582252d4-942f-49a6-9c1c-51d96d7bea1a","resolution":{"observed_at":"2026-08-06T15:54:43.437675Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.426060Z","title":"Revisiting the fragility of influence functions.Neural Networks, 162:581–588, 2023","venue":null,"work_id":"064cb8bc-9e0f-4def-afe6-4341a83fb826","year":2023},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.486191Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:b895b95bfa10bcb9698b47bca60a15ac4fb28f8a16db40f0c9b31ad023ab5409","observation_id":"3351f825-c4ed-4a02-915f-027a6b061ccd","resolution":{"observed_at":"2026-08-06T15:54:43.429039Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.417044Z","title":"A bayesian approach to analysing training data attribution in deep learning.Advances in Neural Information Processing Systems, 36, 2024","venue":null,"work_id":"33993a4f-18f7-4a24-96f9-d60664762c15","year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.489122Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:ee312e5961eff9aecc942357d5e34b5ad6a30779b864d618575302e66b54a259","observation_id":"8db51855-a26a-4a04-b7e1-a5f9540f5094","resolution":{"observed_at":"2026-08-06T15:54:43.420383Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.407605Z","title":"Gradient-based hyperparameter optimization through reversible learning","venue":null,"work_id":"f4f42247-d7cf-4cdf-b86d-9a162a63836c","year":2015},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.491872Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:2f73884d26cbd52984f57493911c9f0fc6912c044f06261b0482f4cf62b7f9dd","observation_id":"b54c5ed8-d54b-459f-b5b4-359823aef38f","resolution":{"observed_at":"2026-08-06T15:54:43.410793Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.398846Z","title":"A scalable laplace approximation for neural networks","venue":null,"work_id":"5f788d1d-7c41-429b-b3d3-4f55896b7e1c","year":2018},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.494950Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:c4de39d01da15f8c30d486aba172de7fc075f776d945fb33cf86023953b014f0","observation_id":"f5ec10a4-e60b-4072-9e76-77d85d412286","resolution":{"observed_at":"2026-08-06T15:54:43.401877Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.389843Z","title":"Lecture 6.5-rmsprop: Divide the gradient by a running average of its recent magnitude.COURSERA: Neural networks for machine learning, 4(2):26, 2012","venue":null,"work_id":"b7be7c8e-502e-4116-ad35-aefa2e2a51f0","year":2012},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.497778Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:627190920dc5ef813cbe6c6fe79457284d4fa5aa79179dcf84796e3d1c476d51","observation_id":"2ed6a9ac-2654-476d-9064-bca4a35c4e7a","resolution":{"observed_at":"2026-08-06T15:54:43.393195Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1412.6980","last_updated":"2017-01-30T01:27:54Z","snapshot_observed_at":"2026-08-17T19:26:44.032537Z","submitted_at":"2014-12-22T13:54:29Z","title":"Adam: A Method for Stochastic Optimization","version":9},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1412.6980","snapshot_observed_at":"2026-08-06T15:54:42.500682Z","title":"Adam: A method for stochastic optimization.arXiv preprint arXiv:1412.6980, 2014","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.500682Z"},"links":{"cited_paper":"/paper/1412.6980","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:188d3e7b6644e26f37e13c3339a7d40b6f2e51964531e9f5f8e2970c5036177d","observation_id":"49279a8a-3ab9-4859-a25e-68729dd64955","resolution":{"observed_at":"2026-08-06T15:54:42.500682Z","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":"2026-08-06T15:54:43.380565Z","title":"A stochastic quasi- newton method for large-scale optimization.SIAM Journal on Optimization, 26(2):1008–1031, 2016","venue":null,"work_id":"579205d6-aeb9-474a-8a41-560795de341a","year":2016},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.503520Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:aeaf5eaa9163ca08a87bd08872c17c17beffa539967a0b8bb0f013e96377e32b","observation_id":"0409a97e-1e0f-4757-a340-bb92354897e2","resolution":{"observed_at":"2026-08-06T15:54:43.383722Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.371449Z","title":"Neural network training dynamics, 2021","venue":null,"work_id":"75be5e9e-195d-4fac-8cc0-23041f246a5a","year":2021},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.506362Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:4af1f9942de87a87a3ffa6c93056cd7a82428edf9846eea056ba393274598809","observation_id":"768c333f-7edb-4b41-b6f6-d87283059732","resolution":{"observed_at":"2026-08-06T15:54:43.374610Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.509630Z","title":"Fast exact multiplication by the hessian.Neural computation, 6(1): 147–160, 1994","venue":null,"work_id":null,"year":1994},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.509630Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:7b7e1cf73d69516e905ab5c7b3454c6b6c6770d5a54a65278fd83ffd911d9caf","observation_id":"e2072b0a-2aee-4e51-b016-9d9684654d17","resolution":{"observed_at":"2026-08-06T15:54:42.509630Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.512599Z","title":"Methods of conjugate gradients for solving linear systems.Journal of research of the National Bureau of Standards, 49(6):409–436, 1952","venue":null,"work_id":null,"year":1952},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.512599Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:564f951a0f09e26e67e2f2672253aadb0dff122d8ee502c8d1e2832a71854ace","observation_id":"ec838b8c-5e08-40c6-9bbe-acd5f3829937","resolution":{"observed_at":"2026-08-06T15:54:42.512599Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.515613Z","title":"Practical bayesian optimization of machine learning algorithms.Advances in neural information processing systems, 25, 2012","venue":null,"work_id":null,"year":2012},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.515613Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:06dc5ac4ecf92827cd16c3840f37052289b2c0f69e69ed9a1abee868b5d4e2b4","observation_id":"a4960d76-45ca-4473-9b05-90125f4c9f17","resolution":{"observed_at":"2026-08-06T15:54:42.515613Z","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":"2026-08-06T15:54:43.346861Z","title":"Influential observations in linear regression.Journal of the American Statistical Association, 74(365):169–174, 1979","venue":null,"work_id":"8e2691dc-d46a-4fef-90c6-b56dc4b06ae4","year":1979},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.518603Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:50bdd66619a9f81c16bd6b9cce57c0f91d0bebe7abf0142c7ee1de45a09fa540","observation_id":"268d1fdf-ea65-4bf8-b959-9c64d6d60c11","resolution":{"observed_at":"2026-08-06T15:54:43.350090Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.521489Z","title":"Fast curvature matrix-vector products for second-order gradient descent","venue":null,"work_id":null,"year":2002},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.521489Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:6f0403009cdc22cd72a2f9d315a0ba21f96c7207f531ad1b0beb7e2009877dda","observation_id":"d1e1200e-771c-4296-a4f9-e689062bcf5c","resolution":{"observed_at":"2026-08-06T15:54:42.521489Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.524379Z","title":"New insights and perspectives on the natural gradient method.Journal of Machine Learning Research, 21(146):1–76, 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.524379Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:735b0c7594fe3c54d114432d86a43505a8849fe91105390b3ff340273e7b35d5","observation_id":"72de3eef-ebbe-4bb3-a718-0929199b294f","resolution":{"observed_at":"2026-08-06T15:54:42.524379Z","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":"2026-08-06T15:54:43.327833Z","title":"The mirrored influence hypothesis: Efficient data influence estimation by harnessing forward passes","venue":null,"work_id":"21e36836-377c-4e87-8345-7cc55722f490","year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.527272Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:a97e1a832356fb1f8d4f930027b26c5b543306875deda7bbba5369b0d1f037fa","observation_id":"98ccf434-3067-4fc0-92f1-60c95314b9fd","resolution":{"observed_at":"2026-08-06T15:54:43.330962Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.319091Z","title":"explainable ai?","venue":null,"work_id":"96e931e5-94bd-4db6-b068-234fdf38ac44","year":2020},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.530341Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:d0cbab707efb9bfbcebb99f9a76f85c4819cba7ed37a602368d0b31cc3e05a5e","observation_id":"b4dff0ac-a55d-4760-a471-aa6820ccfe20","resolution":{"observed_at":"2026-08-06T15:54:43.322075Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.309998Z","title":"Training data influence analysis and estimation: A survey.Machine Learning, 113(5):2351–2403, 2024","venue":null,"work_id":"6d6bc597-c54c-456f-9561-6614fcdac4c6","year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":58,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.533583Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:dcbfe924f3d39219ee0ca2fa0b68672ba52b1a5c4c8c99aaa572b105049b675d","observation_id":"07028868-f55a-4fbc-8a5c-cea027c5782f","resolution":{"observed_at":"2026-08-06T15:54:43.313462Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.301002Z","title":"Combining feature and instance attribution to detect artifacts","venue":null,"work_id":"9af4da7a-904c-4969-8c46-065ba7a230ff","year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":59,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.536546Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:9c873d6d57545cd34be36ec643b448dd1eb393a0dc6ba67f7255d11e26ce4e1f","observation_id":"4bfa9273-c26a-4a25-84ac-6c388f3aa229","resolution":{"observed_at":"2026-08-06T15:54:43.304169Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":"2409.16978","doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.867943Z","title":"Towards user-focused research in training data attribution for human-centered explainable ai.arXiv preprint arXiv:2409.16978, 2024","venue":null,"work_id":"44dee2b5-6e18-419d-bdc6-65737745b9c5","year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":60,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.539510Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:42a09fa928fe9aa2278edc2b95872f6c8c0e0eacbd01c6a94558a4bdf06a4eec","observation_id":"1dd70596-96ba-49c1-bb20-7fce37ae84df","resolution":{"observed_at":"2026-08-06T15:54:42.872716Z","resolver_source":"raw_fallback","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2209.09125","last_updated":"2022-09-16T16:59:36Z","snapshot_observed_at":"2026-08-19T20:22:43.980924Z","submitted_at":"2022-09-16T16:59:36Z","title":"Operationalizing Machine Learning: An Interview Study","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2209.09125","snapshot_observed_at":"2026-08-06T15:54:42.542396Z","title":"Opera- tionalizing machine learning: An interview study.arXiv preprint arXiv:2209.09125, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":61,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.542396Z"},"links":{"cited_paper":"/paper/2209.09125","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:c9aea3030be3aa55b8a17d74990354da2693d78f66474d317205f76c3035b915","observation_id":"3cf48b76-7d20-4929-9b13-2142d4f328aa","resolution":{"observed_at":"2026-08-06T15:54:42.542396Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.545601Z","title":"SIAM, 2003","venue":null,"work_id":null,"year":2003},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":62,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.545601Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:af535b2c62e03190c80b3b729efee08070cb23b0eeb24ff31ae81946c04a51ce","observation_id":"dce471f9-58cf-45fe-8bad-f0dea4dfddb3","resolution":{"observed_at":"2026-08-06T15:54:42.545601Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.00902","last_updated":"2024-03-13T14:27:46Z","snapshot_observed_at":"2026-08-19T23:31:11.622376Z","submitted_at":"2023-10-02T04:59:19Z","title":"DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.00902","snapshot_observed_at":"2026-08-06T15:54:42.548527Z","title":"Datainf: Efficiently estimating data influence in lora-tuned llms and diffusion models.arXiv preprint arXiv:2310.00902, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":63,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.548527Z"},"links":{"cited_paper":"/paper/2310.00902","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:c2b81ade291eff422932f9de170ed2a6ec6a6e248456de247f55296af7637fae","observation_id":"5808b216-d8c7-422f-80df-41426c8e9ac0","resolution":{"observed_at":"2026-08-06T15:54:42.548527Z","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":"2026-08-06T15:54:43.285640Z","title":"Function minimization by conjugate gradients.The computer journal, 7(2):149–154, 1964","venue":null,"work_id":"31dd34d7-9b42-463c-9545-ec99e62a8994","year":1964},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":64,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.551601Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:aab2cac18fc69d875a22fa1903f429683ba017c3a2a9fe22fce809a46fff1923","observation_id":"32d83fbc-baa6-4e59-8868-9a33d2ea6b23","resolution":{"observed_at":"2026-08-06T15:54:43.289539Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.554519Z","title":"Deep learning via hessian-free optimization","venue":null,"work_id":null,"year":2010},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":65,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.554519Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:d449d5365ab379f36d52f1d0d7ed21db61cc355da39ededa0e174c2a6840d056","observation_id":"5e65870e-911e-4546-a0f9-2fc0e3b48eae","resolution":{"observed_at":"2026-08-06T15:54:42.554519Z","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":"2026-08-06T15:54:43.271348Z","title":"dattri: A library for efficient data attribution","venue":null,"work_id":"5453040a-703b-4257-afed-161d6ca35d8e","year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":66,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.557571Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:35a50c2464ec75fba826a7e25572774b088ce0561ac8b46b5d306ea9f28de94a","observation_id":"487b0c0e-cef0-44bb-a471-4da95caa44cf","resolution":{"observed_at":"2026-08-06T15:54:43.274349Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2409.19998","last_updated":"2024-12-19T19:33:43Z","snapshot_observed_at":"2026-08-16T13:14:15.528834Z","submitted_at":"2024-09-30T06:50:18Z","title":"Do Influence Functions Work on Large Language Models?","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.19998","snapshot_observed_at":"2026-08-06T15:54:42.560333Z","title":"Do influence functions work on large language models?arXiv preprint arXiv:2409.19998, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":67,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.560333Z"},"links":{"cited_paper":"/paper/2409.19998","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:e0dda76898e39916369e284124f8c65230969e7a46d6d5533c8558f3b2a6f504","observation_id":"ce782d40-482c-4e36-966c-f5ff0fe9593b","resolution":{"observed_at":"2026-08-06T15:54:42.560333Z","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":"2026-08-06T15:54:43.262464Z","title":"Fletcher and M","venue":null,"work_id":"cc18eca0-3075-4d94-8ba3-eee2e3f51e1f","year":1963},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":68,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.563589Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:01bb7edeba5650a6d596f991e3fcbb658d950a7043ac53021e31fdf8d3b252c2","observation_id":"ba357036-4f6a-4eeb-818c-db1f2465bf5e","resolution":{"observed_at":"2026-08-06T15:54:43.265478Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.253685Z","title":"A class of methods for solving nonlinear simultaneous equations.Mathe- matics of computation, 19(92):577–593, 1965","venue":null,"work_id":"cbfb156d-ade9-486a-9e37-6febb01da174","year":1965},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":69,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.566360Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:8a27e99ef06c409d54d397dcaa0340407b010efef27b35ae9341155f6c8ae994","observation_id":"10931fc8-fca5-4c55-a9c9-64f36005b2b8","resolution":{"observed_at":"2026-08-06T15:54:43.256893Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.244627Z","title":"On the limited memory bfgs method for large scale optimiza- tion.Mathematical programming, 45(1):503–528, 1989","venue":null,"work_id":"9c77e92b-8581-4163-aa0c-756334ac5ca2","year":1989},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":70,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.569744Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:41d5def2ae5792858d1bb5c99de3346098245ba2cda6e6e4e4d3d8bb28e014bd","observation_id":"6c51745d-3611-4383-bc45-80a2a6b026fe","resolution":{"observed_at":"2026-08-06T15:54:43.247860Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.235840Z","title":"Updating quasi-newton matrices with limited storage.Mathematics of compu- tation, 35(151):773–782, 1980","venue":null,"work_id":"c1f9c508-39f5-435f-acb8-ab3ed92782df","year":1980},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":71,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.572489Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:25dd247de217a6a3f707cbd638d4456f8934d67ebf7c0d288d3c73e6286f265a","observation_id":"3bbeb0b7-9cae-4e33-acdd-657de4e37b6c","resolution":{"observed_at":"2026-08-06T15:54:43.238945Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.575338Z","title":"Natural gradient works efficiently in learning.Neural computation, 10(2): 251–276, 1998","venue":null,"work_id":null,"year":1998},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":72,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.575338Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:3a97c5dde99aca30b31f979ec3b9369632af459aa14101fa7feac0ca70459845","observation_id":"1ed5530f-1e6b-410d-8952-f0292d87194c","resolution":{"observed_at":"2026-08-06T15:54:42.575338Z","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":"2026-08-06T15:54:43.221972Z","title":"Noisy natural gradient as variational inference","venue":null,"work_id":"efcb3948-41a8-4d54-9e02-3ab1122fbe39","year":2018},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":73,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.578219Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:77de8656593367247dd4b66753dc7207e00058b93d3829b60154e88e44c223a9","observation_id":"8d9e2584-aa94-448a-b2cb-53d25cf1c652","resolution":{"observed_at":"2026-08-06T15:54:43.224992Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.213128Z","title":"Learning recurrent neural networks with hessian-free optimization","venue":null,"work_id":"bfa2ab7e-e454-4db6-984f-34741788060d","year":2011},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":74,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.581459Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:0406b14b6fbe53e406a43192b2a6db5293d78298c024ce9bd07793e745af314d","observation_id":"054ce2ad-437b-4aa1-a23b-e10d06a93944","resolution":{"observed_at":"2026-08-06T15:54:43.216326Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.204098Z","title":"PhD thesis, University of Toronto (Canada), 2025","venue":null,"work_id":"5304201e-d025-4a6a-b2f2-46c20e2dbcf2","year":2025},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":75,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.584286Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:916cbef8df918eb0981ed38d13a5bd23977e34831e51498ac1d9ad16e23d021b","observation_id":"8b99613a-8a67-4038-be09-9a31e422ed04","resolution":{"observed_at":"2026-08-06T15:54:43.207509Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.587117Z","title":"Uci machine learning repository","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":76,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.587117Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:7928bf97c405b795fc08b1c75911b3d5abe327f2a8bb738c40db51b73343873a","observation_id":"afa01aab-6239-4365-b66a-3bff9b2f3690","resolution":{"observed_at":"2026-08-06T15:54:42.587117Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.589958Z","title":"Learning multiple layers of features from tiny images, 2009","venue":null,"work_id":null,"year":2009},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":77,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.589958Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:ba394a346a9537e20ce442b321b25af62087c3230bad4f9c599d29d726b7b0ac","observation_id":"c1981cda-76a0-4aee-85c7-aa297451a903","resolution":{"observed_at":"2026-08-06T15:54:42.589958Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.592807Z","title":"Deep residual learning for image recognition","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":78,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.592807Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:82e297b398cab2d44268626f6870943d7f3ebfdf37306a3b84f1f98e93735021","observation_id":"5896e329-0f71-4549-a2e5-4ddd3f866730","resolution":{"observed_at":"2026-08-06T15:54:42.592807Z","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":"2026-08-06T15:54:43.179835Z","title":"MNIST handwritten digit database.ATT Labs, 2, 2010","venue":null,"work_id":"88fbc70a-b595-45b8-8d0c-8674c64b5151","year":2010},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":79,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.595594Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:a81c47c4c34483abdc76a814800e0de720f92bbcdbcd7b2591c95163247cac32","observation_id":"5bc8f4e6-38fa-4aef-9edf-dcec7e4d8f7c","resolution":{"observed_at":"2026-08-06T15:54:43.182894Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1708.07747","last_updated":"2017-09-15T21:29:49Z","snapshot_observed_at":"2026-08-13T15:13:33.081929Z","submitted_at":"2017-08-25T14:01:29Z","title":"Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1708.07747","snapshot_observed_at":"2026-08-06T15:54:42.598525Z","title":"Fashion-MNIST: A novel image dataset for benchmarking machine learning algorithms.arXiv preprint arXiv:1708.07747, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":80,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.598525Z"},"links":{"cited_paper":"/paper/1708.07747","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:3632db35bdc7f0d2eb34197c550f43cee67fd11dc66ab6e58337e8a56901d404","observation_id":"60c5bb24-f739-4e94-8b0b-4f8d8576065c","resolution":{"observed_at":"2026-08-06T15:54:42.598525Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.601542Z","title":"Language models are unsupervised multitask learners.OpenAI blog, 1(8):9, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":81,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.601542Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:8deb046bc7145fc693a66fdf9c6d1583d836a06bc857e1d6b7bd8de94b8d0863","observation_id":"74a75e2d-5aea-4f7c-bfb4-2b4dd0d32430","resolution":{"observed_at":"2026-08-06T15:54:42.601542Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1609.07843","last_updated":"2016-09-26T04:06:13Z","snapshot_observed_at":"2026-08-17T01:46:43.513643Z","submitted_at":"2016-09-26T04:06:13Z","title":"Pointer Sentinel Mixture Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1609.07843","snapshot_observed_at":"2026-08-06T15:54:42.604468Z","title":"Pointer sentinel mixture models.arXiv preprint arXiv:1609.07843, 2016","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":82,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.604468Z"},"links":{"cited_paper":"/paper/1609.07843","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:6c56a68684bd6f1a259d0dd098b4bbcd5503f9b3da6314eb9b82c1a2ed069351","observation_id":"65b2789b-9e23-4ab3-b632-cb796b4d4e9f","resolution":{"observed_at":"2026-08-06T15:54:42.604468Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.607478Z","title":"Estimating training data influence by tracing gradient descent.Advances in Neural Information Processing Systems, 33:19920–19930, 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":83,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.607478Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:2260ebbdf5400ed3ba3cf6be7fb5fd97f364e8e392ec96f50db39e0a7de5cbef","observation_id":"5bcef89c-72b6-4588-8af6-b5fee3176ff8","resolution":{"observed_at":"2026-08-06T15:54:42.607478Z","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":"2026-08-06T15:54:43.160885Z","title":"A kronecker-factored approximate fisher matrix for con- volution layers","venue":null,"work_id":"1e5565ca-c18f-4fe4-812f-fe68a424dc3f","year":2016},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":84,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.610381Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:8a91327c7a7d67e497fbe4f5912010b17bdd5790d816262fb3dc9c10606d8889","observation_id":"f7647973-228b-4714-85ab-0eb6de66c626","resolution":{"observed_at":"2026-08-06T15:54:43.163934Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2111.04683","last_updated":"2021-11-08T18:00:06Z","snapshot_observed_at":"2026-08-18T01:33:13.818501Z","submitted_at":"2021-11-08T18:00:06Z","title":"Revisiting Methods for Finding Influential Examples","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2111.04683","snapshot_observed_at":"2026-08-06T15:54:42.613128Z","title":"Revisiting methods for finding influential examples.arXiv preprint arXiv:2111.04683, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":85,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.613128Z"},"links":{"cited_paper":"/paper/2111.04683","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:b10837f502acfd3b8203c1954d5debce4b536a4e7903371f4f4bece3dbff02e2","observation_id":"c371e076-b4bd-4247-8f67-8d00806ec6bb","resolution":{"observed_at":"2026-08-06T15:54:42.613128Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2409.18153","last_updated":"2025-01-08T22:20:36Z","snapshot_observed_at":"2026-08-22T00:37:39.193084Z","submitted_at":"2024-09-25T20:00:23Z","title":"Most Influential Subset Selection: Challenges, Promises, and Beyond","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.18153","snapshot_observed_at":"2026-08-06T15:54:42.616319Z","title":"Most influential subset selection: Challenges, promises, and beyond.arXiv preprint arXiv:2409.18153, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":86,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.616319Z"},"links":{"cited_paper":"/paper/2409.18153","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:754c2eb693f67577ce4ec98501b35ee51adfcca0ce4707eec4fab44b9fee14d2","observation_id":"6807acdc-5dfb-49e3-89ed-5d8af7014b2a","resolution":{"observed_at":"2026-08-06T15:54:42.616319Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.02745","last_updated":"2024-07-18T18:16:59Z","snapshot_observed_at":"2026-08-16T13:46:07.468251Z","submitted_at":"2024-06-04T20:01:39Z","title":"Measuring Stochastic Data Complexity with Boltzmann Influence Functions","version":2},"cited_work":{"arxiv_id":"2406.02745","doi":null,"metadata_source":"pith","pith_arxiv_id":"2406.02745","snapshot_observed_at":"2026-08-06T15:54:42.724762Z","title":"Measuring Stochastic Data Complexity with Boltzmann Influence Functions","venue":"cs.LG","work_id":"0e200cf1-1b5b-4f99-8bc1-4e4fd2e5ade9","year":2024},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":87,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.619289Z"},"links":{"cited_paper":"/paper/2406.02745","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:cd8f9d7a15190f880db4a3cb23fc6338455839d65e42ce60870eb7fd59a0d03b","observation_id":"d9f3325a-75fa-40b4-b621-ec4b881a521e","resolution":{"observed_at":"2026-08-06T15:54:42.727949Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.622291Z","title":"Neural tangent kernel: Convergence and generalization in neural networks.Advances in neural information processing systems, 31, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":88,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.622291Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:62aed0ba0fe5997938ecc090e1fea5a2c5de9604c3beb25988e67a5c69c09a45","observation_id":"4613b99c-b70c-48ee-9edc-c989f7b6e577","resolution":{"observed_at":"2026-08-06T15:54:42.622291Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.625089Z","title":"Wide neural networks of any depth evolve as linear models under gradient descent.Advances in neural information processing systems, 32, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":89,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.625089Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:27afc31d28b03de9423ead446b15dba1bc59d5ef7acbecde594192bb83e34d22","observation_id":"fcb11521-1b2b-446e-80e4-96249f4a2319","resolution":{"observed_at":"2026-08-06T15:54:42.625089Z","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":"2026-08-06T15:54:43.141619Z","title":"More than a toy: Random matrix models predict how real-world neural representations generalize","venue":null,"work_id":"fd8a36a7-89df-464d-ac7a-2ae8016625aa","year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":90,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.627984Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:e5e72018e1a7a5e473830842407ed9b8343ad72fa5809ba0ff6873e0a00f7932","observation_id":"fc3dffa1-5fda-49df-9d65-3e5c37998716","resolution":{"observed_at":"2026-08-06T15:54:43.144793Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.630788Z","title":"Limitations of the empirical fisher approximation for natural gradient descent.Advances in neural information processing systems, 32, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":91,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.630788Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:a38ca02a39950a46b8f84aec3b26274a307aaeeed9fd8014eb4fb0a73d5b1216","observation_id":"be52fd51-44bb-4dc1-bb8d-8aa39273a47b","resolution":{"observed_at":"2026-08-06T15:54:42.630788Z","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":"2026-08-06T15:54:43.127883Z","title":"Understanding influence functions and datamodels via harmonic analysis","venue":null,"work_id":"8b71403c-d960-4995-993a-3f0c4d4979ca","year":2022},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":92,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.633595Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:e5c63257aeec5190ffc85d7a426f008aa8beebc84a9878f5047ac62f34d334a4","observation_id":"a319feb3-92b7-467b-95f1-a90013c10a1b","resolution":{"observed_at":"2026-08-06T15:54:43.130769Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:43.119762Z","title":"On second-order group influence functions for black-box predictions","venue":null,"work_id":"fd7f32da-e7a9-40aa-89fe-efda028b2504","year":2020},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":93,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.636251Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:596691ba3e311c73ea9ee9af6595b5c48e3dba15f13435cf0ca1a5377f8946f5","observation_id":"c4e652a5-1bc6-461e-b9b6-c6ac7f4f5711","resolution":{"observed_at":"2026-08-06T15:54:43.122611Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.639007Z","title":"Simple and scalable predictive uncertainty estimation using deep ensembles.Advances in neural information processing systems, 30, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":94,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.639007Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:71c6fa94dd1de337238ed6b9db512a92de595c5aba6e50692f63be093af1dd7e","observation_id":"65480876-2f70-491a-8739-82a753a7af8e","resolution":{"observed_at":"2026-08-06T15:54:42.639007Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1803.05407","last_updated":"2019-02-25T14:18:11Z","snapshot_observed_at":"2026-08-14T19:36:09.254749Z","submitted_at":"2018-03-14T17:09:27Z","title":"Averaging Weights Leads to Wider Optima and Better Generalization","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1803.05407","snapshot_observed_at":"2026-08-06T15:54:42.641852Z","title":"Averaging weights leads to wider optima and better generalization.arXiv preprint arXiv:1803.05407, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":95,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.641852Z"},"links":{"cited_paper":"/paper/1803.05407","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:5c2fc703cd17d1e764c81b8bf8ee56a4e332f987a3eb0180276c6b47a0b2d899","observation_id":"ec4ee1e3-cf94-47dc-8cc6-73d3abfc7774","resolution":{"observed_at":"2026-08-06T15:54:42.641852Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.645102Z","title":"A simple baseline for bayesian uncertainty in deep learning.Advances in neural information processing systems, 32, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":96,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.645102Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:eb8db9cad18886f2272499e9c725ec710d9bc120cfc570e76af6b373e2ab60fc","observation_id":"00755643-1459-4244-b762-3f76f7b9d809","resolution":{"observed_at":"2026-08-06T15:54:42.645102Z","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":"2026-08-06T15:54:43.100202Z","title":"Bayesian deep learning and a probabilistic perspective of generalization.Advances in neural information processing systems, 33:4697–4708, 2020","venue":null,"work_id":"349829e5-98b6-40d7-8549-b84d082bb0ba","year":2020},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":97,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.647907Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:61e7c1b22a3521ff9c957e2be61f7c8f20c0af25e319b8cf6c73547ed090445b","observation_id":"40e0d2f1-99cc-4b87-b2d5-0bbd4b48293d","resolution":{"observed_at":"2026-08-06T15:54:43.103479Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-22T06:32:14.747728+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.650854Z","title":"Dropout as a bayesian approximation: Representing model uncertainty in deep learning","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":98,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.650854Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:eda9acb7d8641401c29922bf1290a8023c59ade95cea961baeab4b52280be35d","observation_id":"c34c174d-c71e-44aa-bd2e-ea805dfb79ea","resolution":{"observed_at":"2026-08-06T15:54:42.650854Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2303.08114","last_updated":"2023-03-14T17:47:25Z","snapshot_observed_at":"2026-08-19T14:36:40.621810Z","submitted_at":"2023-03-14T17:47:25Z","title":"Simfluence: Modeling the Influence of Individual Training Examples by Simulating Training Runs","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.08114","snapshot_observed_at":"2026-08-06T15:54:42.653576Z","title":"Simfluence: Modeling the influence of individual training examples by simulating training runs.arXiv preprint arXiv:2303.08114, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":99,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.653576Z"},"links":{"cited_paper":"/paper/2303.08114","citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:d00373379258543458efec373aafdc72107bb13806376ceab6ec4bb4ad7aecd5","observation_id":"0340cbe2-c608-40ce-96a7-7bef73112847","resolution":{"observed_at":"2026-08-06T15:54:42.653576Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-06T15:54:42.656667Z","title":"Statistics and causal inference.Journal of the American statistical Association, 81(396):945–960, 1986","venue":null,"work_id":null,"year":1986},"citing_paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products","version":1},"reference_index":100,"source":"pdf_text","source_observed_at":"2026-08-06T15:54:42.656667Z"},"links":{"citing_paper":"/paper/2507.14740"},"observation_digest":"sha256:2aeaf240ba8cb2cddaadb931fb1f5fcc0b9c46ed09b6cab3dd19287d427f737e","observation_id":"f68bf6c9-c425-4159-83bd-60f312aa30db","resolution":{"observed_at":"2026-08-06T15:54:42.656667Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2507.14740","last_updated":"2025-07-19T20:18:51Z","latest_version":1,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-15T11:57:22.622269Z","submitted_at":"2025-07-19T20:18:51Z","title":"Better Training Data Attribution via Better Inverse Hessian-Vector Products"},"reference_resolution":{"displayed":100,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":67,"verified_exact":3,"verified_fuzzy":30},"total_outbound_references":103},"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":"2026-08-22T06:32:14.747728+00:00","source":"crossref"},{"observed_at":"2026-08-22T06:32:06.552537+00:00","source":"retraction_watch"}],"thesis":"As of 22 August 2026, this Paper Citation Record lists 100 of 103 outbound references and 2 inbound Pith citation observations for arXiv:2507.14740."}