{"as_of":"2026-08-16T13:38:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:cce081fb66941077fd28c403fcd05507d58080bbfb8032a409da30abbf316068","coverage":[{"denominator":82,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":82,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-15T20:48:52.385258Z","state":"measured"},{"denominator":85,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":85,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-16T06:30:59.297886+00:00","state":"measured"},{"denominator":3,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":3,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-03T10:37:04.332008Z","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":"2026-05-18T10:46:17.087522Z","state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"cited_work":{"arxiv_id":"2505.11953","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2505.11953","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Exploring criteria of loss reweighting to enhance llm unlearning","venue":null,"work_id":"8d2616b1-375a-4b0e-9c14-4ab15723ec97","year":2024},"citing_paper":{"arxiv_id":"2510.00761","last_updated":"2026-04-18T07:30:06Z","snapshot_observed_at":"2026-08-16T07:19:23.004984Z","submitted_at":"2025-10-01T10:50:14Z","title":"Downgrade to Upgrade: Optimizer Simplification Enhances Robustness in LLM Unlearning","version":5},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-05-18T10:44:53.516653Z"},"links":{"cited_paper":"/paper/2505.11953","citing_paper":"/paper/2510.00761"},"observation_digest":"sha256:9c017956f704bd46c372085601ff40bc031f622c4567df8fc3aff74911982c70","observation_id":"0cc44bcf-5c31-4223-9f34-610e304e85a8","resolution":{"observed_at":"2026-05-18T10:46:17.090538Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.11953","snapshot_observed_at":"2026-08-03T10:37:04.332008Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2601.09624","last_updated":"2026-07-27T02:03:28Z","snapshot_observed_at":"2026-08-13T11:58:33.794205Z","submitted_at":"2026-01-14T16:55:58Z","title":"A Mechanistic Perspective and Circuit-Guided Difficulty Metric for Unlearning","version":2},"reference_index":50,"source":"arxiv_source","source_observed_at":"2026-08-03T10:37:04.332008Z"},"links":{"cited_paper":"/paper/2505.11953","citing_paper":"/paper/2601.09624"},"observation_digest":"sha256:7fc61e4ac9954c48dfe81c2a80026ee8cf42eb23036b7d07712a8545cfc0200b","observation_id":"7a5e48f0-a056-4a63-ae81-e35adacaea11","resolution":{"observed_at":"2026-08-03T10:37:04.332008Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"cited_work":{"arxiv_id":"2505.11953","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2505.11953","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Exploring criteria of loss reweighting to enhance llm unlearning","venue":null,"work_id":"8d2616b1-375a-4b0e-9c14-4ab15723ec97","year":2024},"citing_paper":{"arxiv_id":"2602.23798","last_updated":"2026-05-14T09:06:42Z","snapshot_observed_at":"2026-08-14T13:52:30.456588Z","submitted_at":"2026-02-27T08:39:36Z","title":"MPU: Towards Secure and Privacy-Preserving Knowledge Unlearning for Large Language Models","version":2},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-05-15T18:47:30.221093Z"},"links":{"cited_paper":"/paper/2505.11953","citing_paper":"/paper/2602.23798"},"observation_digest":"sha256:9d58d3607f632b81ba509329043e3f33303fac72fa6a527c14bc29211e7c3327","observation_id":"0f03e384-e7c7-44ee-9816-29417bfae75a","resolution":{"observed_at":"2026-05-15T18:50:16.816367Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2505.11953/citation-record","integrity":"/paper/2505.11953/integrity","json":"/paper/2505.11953/citation-record.json","paper":"/paper/2505.11953"},"outbound":[{"citation":{"cited_paper":{"arxiv_id":"2303.08774","last_updated":"2024-03-04T06:01:33Z","snapshot_observed_at":"2026-08-07T07:30:12.213965Z","submitted_at":"2023-03-15T17:15:04Z","title":"GPT-4 Technical Report","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.08774","snapshot_observed_at":"2026-08-15T20:48:51.966037Z","title":"L., Almeida, D., Altenschmidt, J., Altman, S., Anadkat, S., et al","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":1,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:51.966037Z"},"links":{"cited_paper":"/paper/2303.08774","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:048ea6c96208fb6407fe17994147af14e8c1952524a375adac5b4a0f96071152","observation_id":"b0e96e82-db50-4628-9a47-e0a3b334f3cc","resolution":{"observed_at":"2026-08-15T20:48:51.966037Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2211.05826","last_updated":"2022-11-10T19:30:08Z","snapshot_observed_at":"2026-08-15T08:26:18.212544Z","submitted_at":"2022-11-10T19:30:08Z","title":"The CRINGE Loss: Learning what language not to model","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2211.05826","snapshot_observed_at":"2026-08-15T20:48:51.973096Z","title":"The cringe loss: Learning what language not to model","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":2,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:51.973096Z"},"links":{"cited_paper":"/paper/2211.05826","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:fa21a4f5bb7828db732d08739feeaeda32a85561561d9720532399d4d4814a81","observation_id":"ed5540a8-6274-48da-9185-0f502965013f","resolution":{"observed_at":"2026-08-15T20:48:51.973096Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.06559","last_updated":"2024-06-02T06:24:38Z","snapshot_observed_at":"2026-08-16T02:45:02.748192Z","submitted_at":"2024-06-02T06:24:38Z","title":"Harnessing Business and Media Insights with Large Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.06559","snapshot_observed_at":"2026-08-15T20:48:51.979531Z","title":"P., Narang, N., Rivers, J., Maksey, R., Guan, L., Barrere, L","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":3,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:51.979531Z"},"links":{"cited_paper":"/paper/2406.06559","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:9898689d10b10ef7c6e9981b7f4cdf52b50b0feaff532a3909db344a9ee42b94","observation_id":"1fdc6dce-ec65-49e7-af28-df532bd9c702","resolution":{"observed_at":"2026-08-15T20:48:51.979531Z","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":"2026-08-15T20:48:51.985322Z","title":"R., Christodorescu, M., Datta, A., Feizi, S., et al","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":4,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:51.985322Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:f32e65e47a1b1a6e5c6f46be7cfedb25c0921aaf2e4dc46af75591637549d714","observation_id":"70e73bfd-1c8f-46ee-97f9-55eb862bb729","resolution":{"observed_at":"2026-08-15T20:48:51.985322Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.12038","last_updated":"2024-08-05T21:48:22Z","snapshot_observed_at":"2026-08-12T23:40:40.477831Z","submitted_at":"2024-06-17T19:11:40Z","title":"Soft Prompting for Unlearning in Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.12038","snapshot_observed_at":"2026-08-15T20:48:51.990243Z","title":"Soft prompting for unlearning in large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":5,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:51.990243Z"},"links":{"cited_paper":"/paper/2406.12038","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:72078e786ffe9287c0d08b41a548141e21dc809638e6c90fa2afd304bfcbe6a0","observation_id":"e8b20ad2-ade6-4bd2-8a0c-153d94d98c77","resolution":{"observed_at":"2026-08-15T20:48:51.990243Z","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":"2026-08-15T20:48:51.996030Z","title":"A., Jia, H., Travers, A., Zhang, B., Lie, D., and Papernot, N","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":6,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:51.996030Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:16761b8e39d87ca23e484fbdd4d831e6c7ce5880cad521996bedbb26f203af00","observation_id":"9bed62e7-ad67-4fe0-808f-ba07902a5e44","resolution":{"observed_at":"2026-08-15T20:48:51.996030Z","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":"2026-08-15T20:48:56.397501Z","title":"and Lipton, Z","venue":null,"work_id":"7a69f6eb-66c2-4f08-b6c2-ce91bab40105","year":2019},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":7,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.002351Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:7932fcf16d76cdf72efcf11ba99b0b69f4d62befa025c1dc781c904a69fbf2d5","observation_id":"54068942-f70c-4506-bf81-c657797bbc80","resolution":{"observed_at":"2026-08-15T20:48:56.403406Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2310.20150","last_updated":"2023-10-31T03:35:59Z","snapshot_observed_at":"2026-08-16T09:18:07.723874Z","submitted_at":"2023-10-31T03:35:59Z","title":"Unlearn What You Want to Forget: Efficient Unlearning for LLMs","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.20150","snapshot_observed_at":"2026-08-15T20:48:52.007225Z","title":"and Yang, D","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":8,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.007225Z"},"links":{"cited_paper":"/paper/2310.20150","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:c50105b3f2bbdd72d3ddd561236fb1ba9533891b537a166422c5b85fd9ba648c","observation_id":"d4afebd5-1ca5-4142-a2b5-404f7bb85d83","resolution":{"observed_at":"2026-08-15T20:48:52.007225Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.12329","last_updated":"2025-06-06T14:08:20Z","snapshot_observed_at":"2026-08-15T10:26:47.802809Z","submitted_at":"2024-06-18T06:54:05Z","title":"Opt-Out: Investigating Entity-Level Unlearning for Large Language Models via Optimal Transport","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.12329","snapshot_observed_at":"2026-08-15T20:48:52.012588Z","title":"Snap: Unlearning selective knowledge in large language models with negative instructions","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":9,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.012588Z"},"links":{"cited_paper":"/paper/2406.12329","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:64d188cfa834c18253723803c3f46d8c2fc0e46324fd551bd57ecbb32a72fb24","observation_id":"cdb92f6a-e8b3-42a4-88f8-b42094883b53","resolution":{"observed_at":"2026-08-15T20:48:52.012588Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.07698","last_updated":"2024-06-11T20:26:26Z","snapshot_observed_at":"2026-08-12T23:45:08.613448Z","submitted_at":"2024-06-11T20:26:26Z","title":"Label Smoothing Improves Machine Unlearning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.07698","snapshot_observed_at":"2026-08-15T20:48:52.017725Z","title":"Label smoothing improves machine unlearning","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":10,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.017725Z"},"links":{"cited_paper":"/paper/2406.07698","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:6b2102d41924e5aba4b6e6a772ce667d22437d45cac6046c3628ecf92929b338","observation_id":"df25fdda-04fa-4277-ba98-24c404c320a3","resolution":{"observed_at":"2026-08-15T20:48:52.017725Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.10052","last_updated":"2024-10-16T11:50:27Z","snapshot_observed_at":"2026-08-14T05:18:29.996151Z","submitted_at":"2024-02-15T16:21:14Z","title":"UNDIAL: Self-Distillation with Adjusted Logits for Robust Unlearning in Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.10052","snapshot_observed_at":"2026-08-15T20:48:52.023120Z","title":"R., Lin, H., Belkin, M., Huerta, R., and Vuli \\'c , I","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":11,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.023120Z"},"links":{"cited_paper":"/paper/2402.10052","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:a3af768db8c05e346707eaadc430ca35c5bcb30013ee916c549989e753d9576d","observation_id":"4cd34e26-f4ad-49e8-9c9e-b7c2d4576933","resolution":{"observed_at":"2026-08-15T20:48:52.023120Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.02238","last_updated":"2023-10-04T05:20:19Z","snapshot_observed_at":"2026-08-13T05:58:13.172138Z","submitted_at":"2023-10-03T17:48:14Z","title":"Who's Harry Potter? Approximate Unlearning in LLMs","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.02238","snapshot_observed_at":"2026-08-15T20:48:52.028093Z","title":"and Russinovich, M","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":12,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.028093Z"},"links":{"cited_paper":"/paper/2310.02238","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:a09af5f3d8d1806ce5acfbcb6cbcfbb8ddb75736c4cc5004e894554653f87f62","observation_id":"b1736397-ba8e-4781-b5d3-7a7f834f1860","resolution":{"observed_at":"2026-08-15T20:48:52.028093Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.01306","last_updated":"2024-11-19T18:12:45Z","snapshot_observed_at":"2026-08-12T21:18:02.964833Z","submitted_at":"2024-02-02T10:53:36Z","title":"KTO: Model Alignment as Prospect Theoretic Optimization","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.01306","snapshot_observed_at":"2026-08-15T20:48:52.033166Z","title":"Kto: Model alignment as prospect theoretic optimization","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":13,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.033166Z"},"links":{"cited_paper":"/paper/2402.01306","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:da0cde2c2c2b94fc0b99ef631624fb5632b83d269a48f89bb9c6d7eee31289aa","observation_id":"170da653-20ce-4608-bff6-16cfa1ac7721","resolution":{"observed_at":"2026-08-15T20:48:52.033166Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.12508","last_updated":"2024-04-04T07:45:38Z","snapshot_observed_at":"2026-08-03T17:30:25.153454Z","submitted_at":"2023-10-19T06:17:17Z","title":"SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.12508","snapshot_observed_at":"2026-08-15T20:48:52.038402Z","title":"Salun: Empowering machine unlearning via gradient-based weight saliency in both image classification and generation","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":14,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.038402Z"},"links":{"cited_paper":"/paper/2310.12508","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:9ac5261745964d01440197c6feacdf8a950e1f9cbc9ba901e67e018a765b1c38","observation_id":"252a7b6e-0cfa-4850-9310-ba0828e6bfd8","resolution":{"observed_at":"2026-08-15T20:48:52.038402Z","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":"2026-08-15T20:48:52.043863Z","title":"Simplicity prevails: Rethinking negative preference optimization for llm unlearning","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":15,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.043863Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:42a2db0a3e2dc0eb1e89a983809257427a332e4ab732c61f8d9ae6e2c48a4aae","observation_id":"e4b24885-3aa2-49c8-8ba0-e79f34461c49","resolution":{"observed_at":"2026-08-15T20:48:52.043863Z","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":"2026-08-15T20:48:56.381119Z","title":"Challenging forgets: Unveiling the worst-case forget sets in machine unlearning","venue":null,"work_id":"875261af-c140-4b01-83af-abf12aefa24d","year":2025},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":16,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.048572Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:60de22190ec2c83429cc9468b5e0ebc85b138c6c3b5d51ea33f9ea587051c014","observation_id":"f86e50c4-2eeb-4776-bdbe-07e4c17e9277","resolution":{"observed_at":"2026-08-15T20:48:56.386366Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2407.10223","last_updated":"2025-03-03T01:21:39Z","snapshot_observed_at":"2026-08-16T13:34:22.588125Z","submitted_at":"2024-07-14T14:26:17Z","title":"On Large Language Model Continual Unlearning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.10223","snapshot_observed_at":"2026-08-15T20:48:52.053059Z","title":"Practical unlearning for large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":17,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.053059Z"},"links":{"cited_paper":"/paper/2407.10223","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:76b0b7c1005ca0185d85e094f8f93f9e260b626a275df6cc5041e5b234b54838","observation_id":"2f980a08-9b0b-4e3c-bede-6f17a1cb7ed6","resolution":{"observed_at":"2026-08-15T20:48:52.053059Z","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":"2026-08-15T20:48:56.364198Z","title":"General data protection regulation","venue":null,"work_id":"d1c33d2a-cb95-49f9-8d08-aaeaddafbdd1","year":2016},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":18,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.058356Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:9c25cc9da0bd0fdc1d7f93bd141309d9ec94e7d1f4507334778fdb8b5a02c782","observation_id":"c0775d86-89fc-48e3-aa32-8bc1eb9a5102","resolution":{"observed_at":"2026-08-15T20:48:56.369912Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:52.063010Z","title":"Eternal sunshine of the spotless net: Selective forgetting in deep networks","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":19,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.063010Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:0e7080128fbff468702bb3f7e18be90ae402bbf4e0a839936924d9b23c2f35fb","observation_id":"6b5a1c6b-4b68-4ee6-b26d-0a15d07871ce","resolution":{"observed_at":"2026-08-15T20:48:52.063010Z","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":"2026-08-15T20:48:56.200603Z","title":"Amnesiac machine learning","venue":null,"work_id":"4df5adc2-3bf1-428d-81ed-370bca7f6ca6","year":2021},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":20,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.067699Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:6746f07d91db9f646f65f46ca3b5eba7854a1a615dbed4721b6a0f6ae1602703","observation_id":"1c839a71-b3d5-419f-b630-31ba89838d89","resolution":{"observed_at":"2026-08-15T20:48:56.265948Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2409.11844","last_updated":"2024-09-18T09:55:48Z","snapshot_observed_at":"2026-08-16T13:17:33.692354Z","submitted_at":"2024-09-18T09:55:48Z","title":"MEOW: MEMOry Supervised LLM Unlearning Via Inverted Facts","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.11844","snapshot_observed_at":"2026-08-15T20:48:52.072799Z","title":"Meow: Memory supervised llm unlearning via inverted facts","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":21,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.072799Z"},"links":{"cited_paper":"/paper/2409.11844","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:4767f443be0b9ef527b6cbc715a0c1c5f80c1973bd609f4232f450c67e7a9fe2","observation_id":"bb3343e2-2404-40ca-a940-c6e4395ba892","resolution":{"observed_at":"2026-08-15T20:48:52.072799Z","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":"2026-08-15T20:48:56.184213Z","title":"Aligning ai with shared human values","venue":null,"work_id":"ce21dcc2-65b3-48a6-804b-55db58674e15","year":2021},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":22,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.077617Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:1cba1cde41cc3426dcfc8c26aec6cb94de690763627f5ae8f432413ef7e76121","observation_id":"26cd0d11-20a4-4ecc-96d1-969a580afd46","resolution":{"observed_at":"2026-08-15T20:48:56.190027Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:56.167840Z","title":"Measuring massive multitask language understanding","venue":null,"work_id":"efab745e-ec8c-47c9-8e19-9647ec2113fa","year":2021},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":23,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.083263Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:08faed065ce502e5bf3b64ebbd4992c6551255ed75c1955ae6795d0c0fa440e0","observation_id":"573791f3-2110-46e7-9d5c-8a4223f2711b","resolution":{"observed_at":"2026-08-15T20:48:56.172886Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2404.11045","last_updated":"2025-05-23T00:22:01Z","snapshot_observed_at":"2026-08-13T00:28:37.569505Z","submitted_at":"2024-04-17T03:39:51Z","title":"Offset Unlearning for Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2404.11045","snapshot_observed_at":"2026-08-15T20:48:52.087953Z","title":"Y., Zhou, W., Wang, F., Morstatter, F., Zhang, S., Poon, H., and Chen, M","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":24,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.087953Z"},"links":{"cited_paper":"/paper/2404.11045","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:412c8c1d2d7f858928ff6362c62b206648ad4b0dcf1c20fc7ab3f9b38ac7c28b","observation_id":"8cf17285-7edd-4184-a6c4-a247f6c60afa","resolution":{"observed_at":"2026-08-15T20:48:52.087953Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2401.05561","last_updated":"2024-09-30T10:17:12Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-01-10T22:07:21Z","title":"TrustLLM: Trustworthiness in Large Language Models","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2401.05561","snapshot_observed_at":"2026-08-15T20:48:52.093010Z","title":"Trustllm: Trustworthiness in large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":25,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.093010Z"},"links":{"cited_paper":"/paper/2401.05561","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:13cf04162216f336f162157f5ddb9c27266ab386ee0bb83d5890712f6b289e06","observation_id":"dfc44923-28cf-4991-a3d3-9376abdd6d60","resolution":{"observed_at":"2026-08-15T20:48:52.093010Z","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":"2026-08-15T20:48:55.889769Z","title":"Robust generalization against photon-limited corruptions via worst-case sharpness minimization","venue":null,"work_id":"ce31b620-a443-4a6f-a4d7-727e0f60593e","year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":26,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.097988Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:71d8202894c0c631c7d95bc38911086b2bdd06614cec8e84c49d1d404c3acb2d","observation_id":"4f8ea608-3118-475a-b959-c79210eb0b23","resolution":{"observed_at":"2026-08-15T20:48:55.998501Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2212.04089","last_updated":"2023-03-31T15:27:01Z","snapshot_observed_at":"2026-08-13T03:27:01.609831Z","submitted_at":"2022-12-08T05:50:53Z","title":"Editing Models with Task Arithmetic","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2212.04089","snapshot_observed_at":"2026-08-15T20:48:52.102985Z","title":"T., Wortsman, M., Gururangan, S., Schmidt, L., Hajishirzi, H., and Farhadi, A","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":27,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.102985Z"},"links":{"cited_paper":"/paper/2212.04089","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:5b1a5e0650606554c52bb9e097d7829fd5528b34257302a93b554e41f9de4d18","observation_id":"a079ae72-6ac9-44e4-9ed2-7fbbb2e88660","resolution":{"observed_at":"2026-08-15T20:48:52.102985Z","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":"2026-08-15T20:48:52.108195Z","title":"A., Chaudhuri, K., and Zou, J","venue":null,"work_id":null,"year":2008},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":28,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.108195Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:186d4dd38c806d63553a66c132338ef5cfc1f870651b896361d31a032304911f","observation_id":"93c4af8a-86fa-43fb-ae79-bc5a7d622dc7","resolution":{"observed_at":"2026-08-15T20:48:52.108195Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2210.01504","last_updated":"2022-12-19T11:52:40Z","snapshot_observed_at":"2026-08-13T14:13:28.459055Z","submitted_at":"2022-10-04T10:18:11Z","title":"Knowledge Unlearning for Mitigating Privacy Risks in Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2210.01504","snapshot_observed_at":"2026-08-15T20:48:52.113276Z","title":"Knowledge unlearning for mitigating privacy risks in language models","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":29,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.113276Z"},"links":{"cited_paper":"/paper/2210.01504","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:8ab69250a157c4909ca71b434797337810aa1727ceadb15f9f85bc549c38cf44","observation_id":"e91cd1eb-2e60-42cb-b331-65860e447994","resolution":{"observed_at":"2026-08-15T20:48:52.113276Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.08607","last_updated":"2024-06-12T19:26:35Z","snapshot_observed_at":"2026-08-15T17:59:23.813964Z","submitted_at":"2024-06-12T19:26:35Z","title":"Reversing the Forget-Retain Objectives: An Efficient LLM Unlearning Framework from Logit Difference","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.08607","snapshot_observed_at":"2026-08-15T20:48:52.118339Z","title":"R., Liu, S., and Chang, S","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":30,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.118339Z"},"links":{"cited_paper":"/paper/2406.08607","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:84b4ba8a67ea74c17d28ca4ae88f09112ee4eb57b3de1ab308ad91873c12b478","observation_id":"ab61ec58-ff85-439c-9251-29968fcf8662","resolution":{"observed_at":"2026-08-15T20:48:52.118339Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2404.18239","last_updated":"2024-06-24T20:24:53Z","snapshot_observed_at":"2026-08-15T11:38:10.024463Z","submitted_at":"2024-04-28T16:31:32Z","title":"SOUL: Unlocking the Power of Second-Order Optimization for LLM Unlearning","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2404.18239","snapshot_observed_at":"2026-08-15T20:48:52.123457Z","title":"Soul: Unlocking the power of second-order optimization for llm unlearning","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":31,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.123457Z"},"links":{"cited_paper":"/paper/2404.18239","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:389a8363887c0344625806614fb4ae62edca745b2b9d395756f854e174d79542","observation_id":"03420d9e-91fa-46d8-a4e4-f707822a3364","resolution":{"observed_at":"2026-08-15T20:48:52.123457Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.06825","last_updated":"2023-10-10T17:54:58Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-10-10T17:54:58Z","title":"Mistral 7B","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.06825","snapshot_observed_at":"2026-08-15T20:48:52.128868Z","title":"Q., Sablayrolles, A., Mensch, A., Bamford, C., Chaplot, D","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":32,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.128868Z"},"links":{"cited_paper":"/paper/2310.06825","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:908c92d4f0f857ce00b6ecf6bdd613ac696968b30a798961868baa1eec56a959","observation_id":"1d9a7fa3-3cf2-4402-b2f3-4e60a7f72fdd","resolution":{"observed_at":"2026-08-15T20:48:52.128868Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.13771","last_updated":"2023-10-20T19:14:59Z","snapshot_observed_at":"2026-08-14T00:47:15.826102Z","submitted_at":"2023-10-20T19:14:59Z","title":"Copyright Violations and Large Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.13771","snapshot_observed_at":"2026-08-15T20:48:52.134010Z","title":"Copyright violations and large language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":33,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.134010Z"},"links":{"cited_paper":"/paper/2310.13771","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:31e2afbe968dd02e856119c910739c8bda3f1242f5a202a52e067b86a3539e70","observation_id":"8043da56-74da-4a80-beae-0912296b3857","resolution":{"observed_at":"2026-08-15T20:48:52.134010Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2403.03218","last_updated":"2024-05-15T19:16:09Z","snapshot_observed_at":"2026-08-15T06:18:34.739211Z","submitted_at":"2024-03-05T18:59:35Z","title":"The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning","version":7},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.03218","snapshot_observed_at":"2026-08-15T20:48:52.139236Z","title":"D., Dombrowski, A.-K., Goel, S., Phan, L., et al","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":34,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.139236Z"},"links":{"cited_paper":"/paper/2403.03218","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:d7231b7c6d1ef2ec866e3e4592ef554df24d7dbf7ed42d4360921b3972736a1d","observation_id":"8ecebf5a-f327-4438-b979-523a7f79e572","resolution":{"observed_at":"2026-08-15T20:48:52.139236Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2311.03191","last_updated":"2024-11-28T13:43:50Z","snapshot_observed_at":"2026-08-15T16:37:22.396444Z","submitted_at":"2023-11-06T15:29:30Z","title":"DeepInception: Hypnotize Large Language Model to Be Jailbreaker","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2311.03191","snapshot_observed_at":"2026-08-15T20:48:52.146628Z","title":"Deepinception: Hypnotize large language model to be jailbreaker","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":35,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.146628Z"},"links":{"cited_paper":"/paper/2311.03191","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:1121814d2611e57297f674293203b29c42f1663f44ec3e68a9097ae8d0207891","observation_id":"23e6ae73-2ab6-4d64-ad35-8d2536643af7","resolution":{"observed_at":"2026-08-15T20:48:52.146628Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2309.05463","last_updated":"2023-09-11T14:01:45Z","snapshot_observed_at":"2026-08-02T22:47:03.212781Z","submitted_at":"2023-09-11T14:01:45Z","title":"Textbooks Are All You Need II: phi-1.5 technical report","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2309.05463","snapshot_observed_at":"2026-08-15T20:48:52.151668Z","title":null,"venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":36,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.151668Z"},"links":{"cited_paper":"/paper/2309.05463","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:2c86f4c36649533934b6c4fb42b94e49390c551f34b41c9be20d81f06d85df2a","observation_id":"4351669e-d637-4754-bae2-64e80208b025","resolution":{"observed_at":"2026-08-15T20:48:52.151668Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.07933","last_updated":"2024-10-31T07:36:39Z","snapshot_observed_at":"2026-08-12T23:44:53.015578Z","submitted_at":"2024-06-12T06:56:20Z","title":"Large Language Model Unlearning via Embedding-Corrupted Prompts","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.07933","snapshot_observed_at":"2026-08-15T20:48:52.157064Z","title":"Y., Wang, Y., Flanigan, J., and Liu, Y","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":37,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.157064Z"},"links":{"cited_paper":"/paper/2406.07933","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:06e7997df8f4269fce483db5ec53157719154c0836ae512ac7c0e94e35f96972","observation_id":"f02edcc1-3470-44ca-9226-1eecde3a6a2c","resolution":{"observed_at":"2026-08-15T20:48:52.157064Z","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":"2026-08-15T20:48:55.862496Z","title":"Model sparsity can simplify machine unlearning","venue":null,"work_id":"efdd9e5a-4c58-4da1-abbb-73454a0a4f2c","year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":38,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.162275Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:227169dd4537d6b8aafce679804471727d6570e8ae53890ba8000779eca346a1","observation_id":"4883edd6-7f63-4269-ab28-3bb16fb549d2","resolution":{"observed_at":"2026-08-15T20:48:55.867770Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2304.08733","last_updated":"2025-02-18T06:20:21Z","snapshot_observed_at":"2026-08-13T12:00:27.978026Z","submitted_at":"2023-04-18T05:09:07Z","title":"Human and AI Perceptual Differences in Image Classification Errors","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2304.08733","snapshot_observed_at":"2026-08-15T20:48:52.167325Z","title":"Do humans and machines have the same eyes? human-machine perceptual differences on image classification","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":39,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.167325Z"},"links":{"cited_paper":"/paper/2304.08733","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:ede513c8be3cde5e25859945ca01c42dd2acf846d344b55b44b886d8f57b1b27","observation_id":"08b86906-2c57-408a-9150-45a035da2bd5","resolution":{"observed_at":"2026-08-15T20:48:52.167325Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.11338","last_updated":"2026-04-19T00:17:32Z","snapshot_observed_at":"2026-07-06T19:03:52.342045Z","submitted_at":"2024-08-21T04:45:12Z","title":"Automatic Dataset Construction (ADC): Sample Collection, Data Curation, and Beyond","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.11338","snapshot_observed_at":"2026-08-15T20:48:52.172418Z","title":"Automatic dataset construction (adc): Sample collection, data curation, and beyond","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":40,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.172418Z"},"links":{"cited_paper":"/paper/2408.11338","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:fae9e3490c25ddd80cfa9892baa9ba611a09432bb3fa75639c6e33a31615cc63","observation_id":"4d4d7584-3d8f-4507-a2eb-7446dc65e71b","resolution":{"observed_at":"2026-08-15T20:48:52.172418Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.08787","last_updated":"2024-12-06T21:39:49Z","snapshot_observed_at":"2026-08-14T03:11:49.519597Z","submitted_at":"2024-02-13T20:51:58Z","title":"Rethinking Machine Unlearning for Large Language Models","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.08787","snapshot_observed_at":"2026-08-15T20:48:52.177693Z","title":"Y., Xu, X., Li, H., et al","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":41,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.177693Z"},"links":{"cited_paper":"/paper/2402.08787","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:71797f08f7307d8490b8f5106c5e1ba2bdb53e65880b2969f097a56c166d7e6d","observation_id":"f51eee70-a9c4-4760-9381-122a9066d5b0","resolution":{"observed_at":"2026-08-15T20:48:52.177693Z","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":"2026-08-15T20:48:55.709487Z","title":"and Guo, H","venue":null,"work_id":"33b05b2d-002e-48e6-9ff8-8f4865fbfb18","year":2020},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":42,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.182875Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:3e6f6a4b71b7c18261a40075c0d69c6ed4d4f04fb4103322908c310c4c448e25","observation_id":"6ff84aad-d28c-45df-a713-ade9675e6338","resolution":{"observed_at":"2026-08-15T20:48:55.815603Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:55.693278Z","title":"C., and Kolter, J","venue":null,"work_id":"ea4cbfa7-c41b-440f-a262-8b72c1c678c7","year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":43,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.189032Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:5e6f27b4564a2961fbf4e392a19ae7ee6371626f4a27f12169db5f686b7d54d6","observation_id":"81c9cac4-2d8a-4294-b5ed-a5e961d032c8","resolution":{"observed_at":"2026-08-15T20:48:55.698208Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2409.13474","last_updated":"2024-12-17T17:45:07Z","snapshot_observed_at":"2026-08-16T13:16:52.832959Z","submitted_at":"2024-09-20T13:05:07Z","title":"Alternate Preference Optimization for Unlearning Factual Knowledge in Large Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.13474","snapshot_observed_at":"2026-08-15T20:48:52.193871Z","title":"Alternate preference optimization for unlearning factual knowledge in large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":44,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.193871Z"},"links":{"cited_paper":"/paper/2409.13474","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:45fd3995ace74d984e2ba03f2e6ffeb03840a8a091c1838425ec1d8cc5a552f8","observation_id":"c68b327b-f72d-49b6-afcb-846d7399035e","resolution":{"observed_at":"2026-08-15T20:48:52.193871Z","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":"2026-08-15T20:48:55.675419Z","title":"More human than human: measuring chatgpt political bias","venue":null,"work_id":"a2198b43-1038-4c25-80a4-2e29020b8726","year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":45,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.199112Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:f6f13f2451878c732e857f90bd0ff9f9f4ed1804557e467468cbf43c50a3d63e","observation_id":"bc546bed-4c9e-48dc-942f-f2208721d758","resolution":{"observed_at":"2026-08-15T20:48:55.682364Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:55.660299Z","title":"K., Shokri, R., and Theodorakopoulos, G","venue":null,"work_id":"4afbc2fb-3f14-4dd6-971e-ad29d3d54b68","year":2021},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":46,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.203987Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:e238dc94d20fec88f48caf07fa791fb6d09dd11923e16d02470d2093f33ff0a6","observation_id":"b8bfa195-ad35-4b26-aad4-654962f48732","resolution":{"observed_at":"2026-08-15T20:48:55.664806Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:52.208704Z","title":"R., and Papernot, N","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":47,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.208704Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:8839b615c4e13c2b96f927f52b809d90bc0a4b491b843132bfd329a965e345be","observation_id":"b991a4fa-e2f6-420f-82ca-8fbe1f437686","resolution":{"observed_at":"2026-08-15T20:48:52.208704Z","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":"2026-08-15T20:48:55.502981Z","title":null,"venue":null,"work_id":"8e123653-4e0a-4b01-ada8-e49fac454c32","year":2018},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":48,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.213857Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:d6176c4a65561e387064897eafc553eb7778e79880dbf9f14d5d443fa004a325","observation_id":"b745cf53-04a3-42ac-a22b-e98c7480121d","resolution":{"observed_at":"2026-08-15T20:48:55.623402Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2309.17410","last_updated":"2023-09-29T17:12:43Z","snapshot_observed_at":"2026-08-14T19:08:40.474747Z","submitted_at":"2023-09-29T17:12:43Z","title":"Can Sensitive Information Be Deleted From LLMs? Objectives for Defending Against Extraction Attacks","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2309.17410","snapshot_observed_at":"2026-08-15T20:48:52.218587Z","title":"Can sensitive information be deleted from llms? objectives for defending against extraction attacks","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":49,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.218587Z"},"links":{"cited_paper":"/paper/2309.17410","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:ebb027f42849f99c8039a0f39cb64e6764dd8ab7cbd1450b41593efd84dd1233","observation_id":"fd7c23ff-234e-4949-9d09-e6d2e3662906","resolution":{"observed_at":"2026-08-15T20:48:52.218587Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.07579","last_updated":"2024-06-06T06:31:08Z","snapshot_observed_at":"2026-08-14T11:32:33.344853Z","submitted_at":"2023-10-11T15:19:31Z","title":"In-Context Unlearning: Language Models as Few Shot Unlearners","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.07579","snapshot_observed_at":"2026-08-15T20:48:52.223636Z","title":"In-context unlearning: Language models as few shot unlearners","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":50,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.223636Z"},"links":{"cited_paper":"/paper/2310.07579","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:272be757c92750def4cd3ba7588613613fb68ed29165a216854ee2943f8f71aa","observation_id":"ee62aa73-4c08-489c-9539-61ef16429c1c","resolution":{"observed_at":"2026-08-15T20:48:52.223636Z","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":"2026-08-15T20:48:55.451071Z","title":"Safety alignment should be made more than just a few tokens deep","venue":null,"work_id":"4c6dbfe3-d8b9-4690-b187-15c619152c1b","year":2025},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":51,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.228857Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:d5aec2e099324d5eb84879c0e7eeb224d47f1efcc6b20fcddc1aa10ccaa5ade8","observation_id":"0c8b85ba-1b83-43c0-9214-fd77a788da85","resolution":{"observed_at":"2026-08-15T20:48:55.456367Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:52.233001Z","title":"D., Ermon, S., and Finn, C","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":52,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.233001Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:90971630f926a1c5ce03aab4384b34b18731b36f1347ad72186a88d468a11774","observation_id":"5ae65a7d-7d16-4f58-a252-19b06078a59d","resolution":{"observed_at":"2026-08-15T20:48:52.233001Z","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":"2026-08-15T20:48:55.422927Z","title":"White-box vs black-box: Bayes optimal strategies for membership inference","venue":null,"work_id":"c2063536-b49e-46de-a8be-75bc81245a67","year":2019},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":53,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.237290Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:31dd871b965dc5d45de0fef37c52630c7cf3b85d280c9aaf2dedc2ba10af94fc","observation_id":"b8f26e21-24a4-492b-9751-150c49c586e0","resolution":{"observed_at":"2026-08-15T20:48:55.429873Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2310.16789","last_updated":"2024-03-09T22:26:06Z","snapshot_observed_at":"2026-08-08T18:07:29.632928Z","submitted_at":"2023-10-25T17:21:23Z","title":"Detecting Pretraining Data from Large Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.16789","snapshot_observed_at":"2026-08-15T20:48:52.243466Z","title":"Detecting pretraining data from large language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":54,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.243466Z"},"links":{"cited_paper":"/paper/2310.16789","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:219c11130b2ced950233f03c700ce642369f082812dd34816a7c2869884ff51e","observation_id":"a2d217d8-12fd-4d22-bc22-b705dbc20bc9","resolution":{"observed_at":"2026-08-15T20:48:52.243466Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.06460","last_updated":"2024-07-14T20:14:02Z","snapshot_observed_at":"2026-08-16T13:35:57.672523Z","submitted_at":"2024-07-08T23:47:29Z","title":"MUSE: Machine Unlearning Six-Way Evaluation for Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.06460","snapshot_observed_at":"2026-08-15T20:48:52.247841Z","title":"A., and Zhang, C","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":55,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.247841Z"},"links":{"cited_paper":"/paper/2407.06460","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:9b102734814ac3324c108609086391f5bbf7e02daee6346cb937ecca80ce6c71","observation_id":"4f4fa29e-1e18-4c2e-9f65-6c0c63f24500","resolution":{"observed_at":"2026-08-15T20:48:52.247841Z","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":"2026-08-15T20:48:55.381740Z","title":"Membership inference attacks against machine learning models","venue":null,"work_id":"b8ac2a01-bba9-4475-b49c-1d84ebbb287b","year":2017},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":56,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.252868Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:605f2704fdd0146b57ec518a029630a06ab6c4630453edc2fa884830d93730b0","observation_id":"72605b93-0467-41d2-8cce-7c8fd35567ee","resolution":{"observed_at":"2026-08-15T20:48:55.412074Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2403.03329","last_updated":"2024-06-11T15:47:39Z","snapshot_observed_at":"2026-08-14T05:15:27.896887Z","submitted_at":"2024-03-05T21:19:06Z","title":"Guardrail Baselines for Unlearning in LLMs","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.03329","snapshot_observed_at":"2026-08-15T20:48:52.258038Z","title":"S., and Smith, V","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":57,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.258038Z"},"links":{"cited_paper":"/paper/2403.03329","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:853ed4078d91cd5d7ee4dd7cad0c9942c2f3b1628fa10258548240f89eb9e0e6","observation_id":"4ef2c200-c407-487e-b8dc-de1381e33833","resolution":{"observed_at":"2026-08-15T20:48:52.258038Z","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":"2026-08-15T20:48:52.264058Z","title":"Unrolling sgd: Understanding factors influencing machine unlearning","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":58,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.264058Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:e6fa216a019d59832d416963b89aad760f1df4de2284ba2338fdcecbd12bf9e5","observation_id":"2fd09203-0cf1-4a54-97ab-ec3933a73302","resolution":{"observed_at":"2026-08-15T20:48:52.264058Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2307.09288","last_updated":"2023-07-19T17:08:59Z","snapshot_observed_at":"2026-08-07T12:56:43.323460Z","submitted_at":"2023-07-18T14:31:57Z","title":"Llama 2: Open Foundation and Fine-Tuned Chat Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2307.09288","snapshot_observed_at":"2026-08-15T20:48:52.268844Z","title":"Llama 2: Open foundation and fine-tuned chat models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":59,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.268844Z"},"links":{"cited_paper":"/paper/2307.09288","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:f4a43198c24dbf630e0154de28bcf4878f6bdd5cc519076022bb1fea7e04fc87","observation_id":"3775d31f-b63d-4440-acf7-457a795ff408","resolution":{"observed_at":"2026-08-15T20:48:52.268844Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.06535","last_updated":"2023-05-11T02:44:29Z","snapshot_observed_at":"2026-08-13T22:38:47.024893Z","submitted_at":"2023-05-11T02:44:29Z","title":"KGA: A General Machine Unlearning Framework Based on Knowledge Gap Alignment","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.06535","snapshot_observed_at":"2026-08-15T20:48:52.273960Z","title":"Kga: A general machine unlearning framework based on knowledge gap alignment","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":60,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.273960Z"},"links":{"cited_paper":"/paper/2305.06535","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:1993ffd150757c1c90a1d7b98fc4baf7a41ac6214353458be1c4e1eba7dfbb50","observation_id":"2482fa1b-1d71-46ba-976c-c96c92c550d8","resolution":{"observed_at":"2026-08-15T20:48:52.273960Z","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":"2026-08-15T20:48:54.041029Z","title":"Learning to augment distributions for out-of-distribution detection","venue":null,"work_id":"7e489e3c-cc20-48ad-a5d7-bb0d879b9c63","year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":61,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.279310Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:10ab67aed94854e837855fe1ce226b845da95f6570d02405e770ca69e38039ed","observation_id":"e37b96ee-adb2-44ca-9985-ec8721456fb8","resolution":{"observed_at":"2026-08-15T20:48:55.235847Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:53.978937Z","title":"A sober look at the robustness of clips to spurious features","venue":null,"work_id":"a0df36c0-8504-40cc-956f-cf9278b7ecef","year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":62,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.284376Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:9a984544b237b2ffbe746803b9f729874b137f3961f2c8f87e9acf0ec78e00ec","observation_id":"1b573b71-b072-4ba8-bc82-bfc4a2b05f21","resolution":{"observed_at":"2026-08-15T20:48:54.015885Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:53.912181Z","title":"Towards effective evaluations and comparison for llm unlearning methods","venue":null,"work_id":"45132b91-469a-4a72-8d93-3e809c0235f2","year":2025},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":63,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.289089Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:3cdded92144ffd71fa751d70fee32d7fc20ba9db01f21ca409bf97e329038399","observation_id":"6a748851-dd49-4c2e-be87-0efa65596dd9","resolution":{"observed_at":"2026-08-15T20:48:53.920634Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:53.895547Z","title":"P., Zhou, Z., Shin, S., Han, B., and Weinberger, K","venue":null,"work_id":"651a1e29-68ea-400d-af28-cf5f55378acc","year":2025},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":64,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.294449Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:5bc7f6169f633672370727be466bf45e671797fffa05a4f23f5f4f532ba1f76c","observation_id":"ac918638-9d4d-4105-9c1f-7a9b68e313d6","resolution":{"observed_at":"2026-08-15T20:48:53.900476Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2212.10560","last_updated":"2023-05-25T23:50:07Z","snapshot_observed_at":"2026-08-15T15:06:42.719366Z","submitted_at":"2022-12-20T18:59:19Z","title":"Self-Instruct: Aligning Language Models with Self-Generated Instructions","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2212.10560","snapshot_observed_at":"2026-08-15T20:48:52.299803Z","title":"A., Khashabi, D., and Hajishirzi, H","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":65,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.299803Z"},"links":{"cited_paper":"/paper/2212.10560","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:3a66efd05653de1905f5fa0cb913767cc97f6a2b959fb65a1bcfea08f749a001","observation_id":"2495d27c-69db-40e8-a88c-27b835b057ee","resolution":{"observed_at":"2026-08-15T20:48:52.299803Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2410.11143","last_updated":"2024-10-14T23:43:33Z","snapshot_observed_at":"2026-08-16T13:09:23.596592Z","submitted_at":"2024-10-14T23:43:33Z","title":"LLM Unlearning via Loss Adjustment with Only Forget Data","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2410.11143","snapshot_observed_at":"2026-08-15T20:48:52.305218Z","title":"Y., Pang, J., Liu, Q., Shah, A","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":66,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.305218Z"},"links":{"cited_paper":"/paper/2410.11143","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:3ac4aaf4766e59fdba199e861da33186a56a47233a2517b20c67778a10788efb","observation_id":"2beb9b70-7e4c-4801-90ba-adcf1702c692","resolution":{"observed_at":"2026-08-15T20:48:52.305218Z","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":"2026-08-15T20:48:53.779448Z","title":"Gru: Mitigating the trade-off between unlearning and retention for large language models","venue":null,"work_id":"c8335c20-8c2f-4c75-b2c1-60e65d0c5b1e","year":2025},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":67,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.310703Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:f16e2efd09fd248041e0ebacd6d01afefa179dc672c02f66969a1226ec36f658","observation_id":"007996e0-d566-4f60-b355-0c6b0f9edb2f","resolution":{"observed_at":"2026-08-15T20:48:53.859784Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2110.12088","last_updated":"2022-03-27T06:51:57Z","snapshot_observed_at":"2026-08-13T17:47:52.014143Z","submitted_at":"2021-10-22T22:42:11Z","title":"Learning with Noisy Labels Revisited: A Study Using Real-World Human Annotations","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2110.12088","snapshot_observed_at":"2026-08-15T20:48:52.315327Z","title":"Learning with noisy labels revisited: A study using real-world human annotations","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":68,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.315327Z"},"links":{"cited_paper":"/paper/2110.12088","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:ba7131c491645ff18afaadf66a839e493d7f0e42962649093fff981cb5b23d60","observation_id":"fe053807-2f55-40cf-93d0-37b6e17f65a1","resolution":{"observed_at":"2026-08-15T20:48:52.315327Z","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":"2026-08-15T20:48:52.320353Z","title":"V., Zhou, D., et al","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":69,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.320353Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:a8f14b7e6f825d174ce53d64db285bffbc44b3c2f22383cdafcd987346f31793","observation_id":"45dac47b-91c6-445f-803e-2d06dbd22b53","resolution":{"observed_at":"2026-08-15T20:48:52.320353Z","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":"2026-08-15T20:48:53.746680Z","title":"Trustworthy graph learning: Reliability, explainability, and privacy protection","venue":null,"work_id":"1a0bd29f-5840-4f37-901a-586adbab0f9d","year":2022},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":70,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.325383Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:29ea87e9977fa9474fc6b218dde7c47d1878f2d0a018f10ffe97b57164dba43c","observation_id":"090be6e3-228c-4136-8ede-31e533e91970","resolution":{"observed_at":"2026-08-15T20:48:53.752114Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:53.722500Z","title":"Adaptive localization of knowledge negation for continual llm unlearning","venue":null,"work_id":"38fd5288-de32-484b-862e-3ef070348e86","year":2025},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":71,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.330319Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:3438b99bbf9623f26eb42a828aeb0cf53ee06b1bafb57949ee880557a7db0ba7","observation_id":"1929c9a2-bdd9-46de-a94f-d1d990e3ead2","resolution":{"observed_at":"2026-08-15T20:48:53.735196Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2310.10683","last_updated":"2024-02-16T19:47:36Z","snapshot_observed_at":"2026-08-14T05:38:48.551002Z","submitted_at":"2023-10-14T00:32:55Z","title":"Large Language Model Unlearning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.10683","snapshot_observed_at":"2026-08-15T20:48:52.335545Z","title":"Large language model unlearning","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":72,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.335545Z"},"links":{"cited_paper":"/paper/2310.10683","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:001711040877c881c8c8001f7ebdfde4be3049859691026c7da9ce6ad89c405c","observation_id":"152f79c7-035f-40d4-aa97-12d24923d436","resolution":{"observed_at":"2026-08-15T20:48:52.335545Z","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":"2026-08-15T20:48:52.340566Z","title":"K., Bindschaedler, V., and Shokri, R","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":73,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.340566Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:947d0a4882756a37f6dc4b5f418b8b9a363a36ceda7553c41758aa5678a34581","observation_id":"1f2530c1-5541-4fdd-add0-60855f50ed0b","resolution":{"observed_at":"2026-08-15T20:48:52.340566Z","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":"2026-08-15T20:48:53.580829Z","title":"Towards safe machine unlearning: a paradigm that mitigates performance degradation","venue":null,"work_id":"b3583aa4-4a2e-4843-86ba-700b38e8d9d4","year":2025},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":74,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.345460Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:e4b53365c1d8818478f132694212877c666cc409db3389dd1eaebe244596bc70","observation_id":"e307fc16-9f0d-48a4-a260-de208759452f","resolution":{"observed_at":"2026-08-15T20:48:53.658573Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:53.458812Z","title":"Unlearning bias in language models by partitioning gradients","venue":null,"work_id":"c857aa02-bc05-44c7-8f12-0e00c7dbf386","year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":75,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.350400Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:0402e9bcdedf0a8f90e9750407e2fc9e261ff722b18731d8b704147e9fe1b61e","observation_id":"7f4eaee7-8935-448a-8906-3868b41ee1dc","resolution":{"observed_at":"2026-08-15T20:48:53.516847Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:53.440761Z","title":"Mind the label shift of augmentation-based graph ood generalization","venue":null,"work_id":"a1fa3427-ac3b-45ac-938a-e5816f42a2f0","year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":76,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.355608Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:f2cbe10d7b227df2ff1519b4dc33994079aa63586611f50168ee15d1352c1a26","observation_id":"a8975dba-6b2e-48d1-bc06-7a15e1b72a5d","resolution":{"observed_at":"2026-08-15T20:48:53.446504Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:53.423680Z","title":"Thought propagation: An analogical approach to complex reasoning with large language models","venue":null,"work_id":"462a6d09-ef20-49ef-bc62-14b453fe01d9","year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":77,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.359693Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:7b5164c3de793b4004926b9f04a03df89f77dbf13449736b0527c38404ddd2c4","observation_id":"9094edfc-01a5-44d0-a5a7-7b1979fa4593","resolution":{"observed_at":"2026-08-15T20:48:53.429007Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2404.05868","last_updated":"2024-10-10T22:00:41Z","snapshot_observed_at":"2026-07-06T17:57:27.510162Z","submitted_at":"2024-04-08T21:05:42Z","title":"Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2404.05868","snapshot_observed_at":"2026-08-15T20:48:52.364613Z","title":"Negative preference optimization: From catastrophic collapse to effective unlearning","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":78,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.364613Z"},"links":{"cited_paper":"/paper/2404.05868","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:200760444a2c4c3e45560aa192786bc49e6b4817eecc473cd619d5b16cfdc448","observation_id":"654235b4-1ce4-4a33-b280-880a981f0c85","resolution":{"observed_at":"2026-08-15T20:48:52.364613Z","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":"2026-08-15T20:48:53.382852Z","title":"Can language models perform robust reasoning in chain-of-thought prompting with noisy rationales? In NeurIPS, 2024","venue":null,"work_id":"ac9f45e8-337d-44c1-bf5b-f86f4708e9e7","year":2024},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":79,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.370627Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:e4e54ce26b8e485027c6a0ec337a44b5400e7705ed5c2ecd71b402d5c613fc9e","observation_id":"51e13328-4318-4f07-8d6b-99cfc0859d22","resolution":{"observed_at":"2026-08-15T20:48:53.403334Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+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":"2026-08-15T20:48:52.375391Z","title":"Landscape of thoughts: Visualizing the reasoning process of large language models","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":80,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.375391Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:8163c1a7e707dc974b5405dfa05be5aa4bbe74f1e7bd886651f6b0c2e23fc9cf","observation_id":"657c8ae0-09d2-464d-a296-5ead5dcb7d28","resolution":{"observed_at":"2026-08-15T20:48:52.375391Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2311.11202","last_updated":"2024-03-24T22:02:47Z","snapshot_observed_at":"2026-08-13T05:22:32.926427Z","submitted_at":"2023-11-19T02:34:12Z","title":"Unmasking and Improving Data Credibility: A Study with Datasets for Training Harmless Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2311.11202","snapshot_observed_at":"2026-08-15T20:48:52.380302Z","title":"Unmasking and improving data credibility: A study with datasets for training harmless language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":81,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.380302Z"},"links":{"cited_paper":"/paper/2311.11202","citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:9215da8d4bfa7dbdbc69d46c650548c277f30c386ab140e398c6f7fbf4bb2065","observation_id":"92ec4895-9e86-48a4-b977-03b29c7b3363","resolution":{"observed_at":"2026-08-15T20:48:52.380302Z","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":"2026-08-15T20:48:52.385258Z","title":"write newline","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning","version":2},"reference_index":82,"source":"arxiv_source","source_observed_at":"2026-08-15T20:48:52.385258Z"},"links":{"citing_paper":"/paper/2505.11953"},"observation_digest":"sha256:1d7eb00076384fd84f873d16802a225b20b255fa8d770bb5d943e798e48db3f3","observation_id":"9b18cdda-426f-44c5-a5d6-84ce83ec5d5d","resolution":{"observed_at":"2026-08-15T20:48:52.385258Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2505.11953","last_updated":"2025-05-28T03:33:24Z","latest_version":2,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-16T02:45:35.232572Z","submitted_at":"2025-05-17T10:41:22Z","title":"Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning"},"reference_resolution":{"displayed":82,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":55,"verified_exact":0,"verified_fuzzy":27},"total_outbound_references":82},"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":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"thesis":"As of 16 August 2026, this Paper Citation Record lists 82 of 82 outbound references and 3 inbound Pith citation observations for arXiv:2505.11953."}