{"as_of":"2026-08-09T08:33:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:e865349017dbf678a47c94d394d8f92fefa0b7c1c17e44e52d438ab3445b2d9c","coverage":[{"denominator":0,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":6,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":6,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-09T06:31:02.800959+00:00","state":"measured"},{"denominator":6,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":6,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-07T10:58:30.941185Z","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":"2026-05-24T02:25:55.507074Z","state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2402.18540","last_updated":"2025-01-17T01:43:21Z","snapshot_observed_at":"2026-08-09T01:10:23.759600Z","submitted_at":"2024-02-28T18:23:49Z","title":"Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates","version":2},"cited_work":{"arxiv_id":"2402.18540","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2402.18540","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Keeping llms aligned after fine-tuning: The crucial role of prompt templates","venue":null,"work_id":"65f0bbf0-8f01-4e41-a653-2c30064b6e82","year":2024},"citing_paper":{"arxiv_id":"2404.09005","last_updated":"2026-05-04T06:38:56Z","snapshot_observed_at":"2026-07-06T17:59:46.583076Z","submitted_at":"2024-04-13T13:18:40Z","title":"Proof-of-Learning with Incentive Security","version":8},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-05-24T02:24:17.188754Z"},"links":{"cited_paper":"/paper/2402.18540","citing_paper":"/paper/2404.09005"},"observation_digest":"sha256:a8831452a2c00b657f1db1634f0ac8d09c5010724e4376cdc97dedb9d444f479","observation_id":"fc30f4cd-f9a0-4d2f-9ac7-58a61d6fcd7a","resolution":{"observed_at":"2026-05-24T02:25:55.510376Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.18540","last_updated":"2025-01-17T01:43:21Z","snapshot_observed_at":"2026-08-09T01:10:23.759600Z","submitted_at":"2024-02-28T18:23:49Z","title":"Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates","version":2},"cited_work":{"arxiv_id":"2402.18540","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2402.18540","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Keeping llms aligned after fine-tuning: The crucial role of prompt templates","venue":null,"work_id":"65f0bbf0-8f01-4e41-a653-2c30064b6e82","year":2024},"citing_paper":{"arxiv_id":"2409.18169","last_updated":"2026-04-23T18:48:49Z","snapshot_observed_at":"2026-07-06T19:22:58.341345Z","submitted_at":"2024-09-26T17:55:22Z","title":"Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey","version":6},"reference_index":103,"source":"pdf_text","source_observed_at":"2026-05-23T20:58:16.237327Z"},"links":{"cited_paper":"/paper/2402.18540","citing_paper":"/paper/2409.18169"},"observation_digest":"sha256:099a7b7a157149e671170f46194f7666f2da71bf6acc8e03a74eaad45e78622e","observation_id":"44891152-4c8f-47ea-b4e1-d598e46f4d77","resolution":{"observed_at":"2026-05-23T20:58:26.413890Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.18540","last_updated":"2025-01-17T01:43:21Z","snapshot_observed_at":"2026-08-09T01:10:23.759600Z","submitted_at":"2024-02-28T18:23:49Z","title":"Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.18540","snapshot_observed_at":"2026-08-07T10:58:30.941185Z","title":"Keeping llms aligned after fine-tuning: The crucial role of prompt templates","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.03850","last_updated":"2025-08-12T10:16:47Z","snapshot_observed_at":"2026-08-08T08:07:29.820388Z","submitted_at":"2025-06-04T11:33:36Z","title":"Vulnerability-Aware Alignment: Mitigating Uneven Forgetting in Harmful Fine-Tuning","version":2},"reference_index":26,"source":"arxiv_source","source_observed_at":"2026-08-07T10:58:30.941185Z"},"links":{"cited_paper":"/paper/2402.18540","citing_paper":"/paper/2506.03850"},"observation_digest":"sha256:4f65488dba02ff80bf1c75821fd27e8cda4fe598966f618764635f42f0fb8840","observation_id":"4fed7e0f-c7a7-4eb8-808c-4691a40cde4a","resolution":{"observed_at":"2026-08-07T10:58:30.941185Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.18540","last_updated":"2025-01-17T01:43:21Z","snapshot_observed_at":"2026-08-09T01:10:23.759600Z","submitted_at":"2024-02-28T18:23:49Z","title":"Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.18540","snapshot_observed_at":"2026-08-06T21:27:34.954267Z","title":"Keep- ing llms aligned after fine-tuning: The crucial role of prompt templates","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.00239","last_updated":"2025-06-30T20:13:49Z","snapshot_observed_at":"2026-08-08T11:00:52.585789Z","submitted_at":"2025-06-30T20:13:49Z","title":"Linearly Decoding Refused Knowledge in Aligned Language Models","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-06T21:27:34.954267Z"},"links":{"cited_paper":"/paper/2402.18540","citing_paper":"/paper/2507.00239"},"observation_digest":"sha256:204e580472c0f2fa1af3c9fc73b809253eeee3718a0cc278752d7e85993a67ed","observation_id":"70fcb23a-e528-4821-9f89-888045f82745","resolution":{"observed_at":"2026-08-06T21:27:34.954267Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.18540","last_updated":"2025-01-17T01:43:21Z","snapshot_observed_at":"2026-08-09T01:10:23.759600Z","submitted_at":"2024-02-28T18:23:49Z","title":"Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates","version":2},"cited_work":{"arxiv_id":"2402.18540","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2402.18540","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Keeping llms aligned after fine-tuning: The crucial role of prompt templates","venue":null,"work_id":"65f0bbf0-8f01-4e41-a653-2c30064b6e82","year":2024},"citing_paper":{"arxiv_id":"2604.16659","last_updated":"2026-04-17T19:28:07Z","snapshot_observed_at":"2026-07-06T23:03:57.199731Z","submitted_at":"2026-04-17T19:28:07Z","title":"Benign Fine-Tuning Breaks Safety Alignment in Audio LLMs","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-05-10T08:01:25.938248Z"},"links":{"cited_paper":"/paper/2402.18540","citing_paper":"/paper/2604.16659"},"observation_digest":"sha256:5c8191b6c6939ba96b27acd433481d64eed8f56b5a7758b6b49394607ebf300e","observation_id":"53cbdb86-343d-4044-9d33-c2070408d855","resolution":{"observed_at":"2026-05-10T08:02:24.920882Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.18540","last_updated":"2025-01-17T01:43:21Z","snapshot_observed_at":"2026-08-09T01:10:23.759600Z","submitted_at":"2024-02-28T18:23:49Z","title":"Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates","version":2},"cited_work":{"arxiv_id":"2402.18540","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2402.18540","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Keeping llms aligned after fine-tuning: The crucial role of prompt templates","venue":null,"work_id":"65f0bbf0-8f01-4e41-a653-2c30064b6e82","year":2024},"citing_paper":{"arxiv_id":"2605.19940","last_updated":"2026-05-19T15:00:06Z","snapshot_observed_at":"2026-07-31T21:52:39.654641Z","submitted_at":"2026-05-19T15:00:06Z","title":"Robotics-Inspired Guardrails for Foundation Models in Socially Sensitive Domains","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-05-20T05:45:02.563352Z"},"links":{"cited_paper":"/paper/2402.18540","citing_paper":"/paper/2605.19940"},"observation_digest":"sha256:8f574dc5a0c3393073699f130d6c45d8b3940bc3dc69005d146ce61e27eccc37","observation_id":"0ba554f1-c066-4cb9-957a-7368f98cdfd3","resolution":{"observed_at":"2026-05-20T05:48:04.442559Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2402.18540/citation-record","integrity":"/paper/2402.18540/integrity","json":"/paper/2402.18540/citation-record.json","paper":"/paper/2402.18540"},"outbound":[],"paper":{"arxiv_id":"2402.18540","last_updated":"2025-01-17T01:43:21Z","latest_version":2,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-09T01:10:23.759600Z","submitted_at":"2024-02-28T18:23:49Z","title":"Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates"},"reference_resolution":{"displayed":0,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":0,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":0},"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":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"thesis":"As of 9 August 2026, this Paper Citation Record lists 0 of 0 outbound references and 6 inbound Pith citation observations for arXiv:2402.18540."}