{"as_of":"2026-08-10T06:47:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:d2ba612c072114e87a099107fef4865c2aaf6354e0f227d03baf653966a73f12","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":27,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":27,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-10T06:31:04.303077+00:00","state":"measured"},{"denominator":27,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":27,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-09T18:15:32.881971Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":1,"source":"arxiv_reference","source_observed_at":"2026-08-05T02:28:24.338817Z","state":"measured"}],"external_citation_measurements":[{"count":2,"observed_at":"2026-08-05T02:28:24.338817Z","source":"arxiv_reference"}],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","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":86,"source":"pdf_text","source_observed_at":"2026-05-23T20:58:16.237327Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2409.18169"},"observation_digest":"sha256:bc4a774d6ad131f335600b6aee028b0d88dbf27880500a9121a392cd669f2632","observation_id":"c9d6c6d8-a600-496a-9fc6-c334deb18f87","resolution":{"observed_at":"2026-05-23T20:58:26.349920Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-09T18:15:32.881971Z","title":"Safety layers in aligned large language models: The key to llm security","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2502.00669","last_updated":"2025-02-02T04:43:35Z","snapshot_observed_at":"2026-08-09T18:06:27.732817Z","submitted_at":"2025-02-02T04:43:35Z","title":"Safety Alignment Depth in Large Language Models: A Markov Chain Perspective","version":1},"reference_index":21,"source":"arxiv_source","source_observed_at":"2026-08-09T18:15:32.881971Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2502.00669"},"observation_digest":"sha256:eac79f34452a4748a6229ce9c2462a9b35cabf367ceaa851a8542e4bb4f0b575","observation_id":"4fb17518-3c4a-4b3e-9f8c-6a4a76fceaf7","resolution":{"observed_at":"2026-08-09T18:15:32.881971Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-09T15:22:32.459148Z","title":"Safety layers in aligned large language models: The key to llm security.https://arxiv.org/abs/2408.17003, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2502.01436","last_updated":"2025-12-19T08:17:09Z","snapshot_observed_at":"2026-08-10T02:18:24.339372Z","submitted_at":"2025-02-03T15:19:28Z","title":"Towards Safer Chatbots: Automated Policy Compliance Evaluation of Custom GPTs","version":3},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-09T15:22:32.459148Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2502.01436"},"observation_digest":"sha256:b03e9d276bf9e633e5538bda1c6354c8349181bd27b4f42187571c5604e0760f","observation_id":"3d8041f8-1dac-40b8-b1df-e5b28aadaac6","resolution":{"observed_at":"2026-08-09T15:22:32.459148Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-07T23:08:02.302336Z","title":"Safety layers in aligned large language models: The key to llm security","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2502.09674","last_updated":"2025-05-27T08:40:42Z","snapshot_observed_at":"2026-08-09T02:24:15.276351Z","submitted_at":"2025-02-13T06:39:22Z","title":"The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Analysis of Orthogonal Safety Directions","version":4},"reference_index":26,"source":"arxiv_source","source_observed_at":"2026-08-07T23:08:02.302336Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2502.09674"},"observation_digest":"sha256:5d3198ec730ef5d75b5962a4cabb717b6f94c3e8731b94589250e0720dd1852a","observation_id":"9a175d9f-deb9-4e8c-8ceb-07f6bee309d0","resolution":{"observed_at":"2026-08-07T23:08:02.302336Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-07T14:15:19.536862Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2505.19670","last_updated":"2025-05-26T08:25:25Z","snapshot_observed_at":"2026-08-09T12:47:51.547477Z","submitted_at":"2025-05-26T08:25:25Z","title":"Reshaping Representation Space to Balance the Safety and Over-rejection in Large Audio Language Models","version":1},"reference_index":15,"source":"arxiv_source","source_observed_at":"2026-08-07T14:15:19.536862Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2505.19670"},"observation_digest":"sha256:63bee8f4401ac77954d4ad08b02fec62cfc2e0cf82989ea0d7362c465f78733d","observation_id":"7dd8fe59-9419-43ff-9ff5-03d46611f82f","resolution":{"observed_at":"2026-08-07T14:15:19.536862Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-06T19:57:43.958573Z","title":"Safety layers in aligned large language models: The key to llm security","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.04250","last_updated":"2025-07-06T05:47:04Z","snapshot_observed_at":"2026-08-09T00:28:34.913528Z","submitted_at":"2025-07-06T05:47:04Z","title":"Just Enough Shifts: Mitigating Over-Refusal in Aligned Language Models with Targeted Representation Fine-Tuning","version":1},"reference_index":12,"source":"arxiv_source","source_observed_at":"2026-08-06T19:57:43.958573Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2507.04250"},"observation_digest":"sha256:9c59ee546af242533687a1c177e570445dfc4f4d9699254a422ab7f58f48731c","observation_id":"7cc02672-ba13-4cb4-8174-ab22e8cf3187","resolution":{"observed_at":"2026-08-06T19:57:43.958573Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-06T15:18:11.656139Z","title":"Safety layers of aligned large language models: The key to llm security","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.16372","last_updated":"2025-07-22T09:15:11Z","snapshot_observed_at":"2026-08-08T20:51:05.150695Z","submitted_at":"2025-07-22T09:15:11Z","title":"Depth Gives a False Sense of Privacy: LLM Internal States Inversion","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-06T15:18:11.656139Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2507.16372"},"observation_digest":"sha256:0bfcc0e16d68155c5c7ce4935e49fd985a44945826c2b9c81d0c525d742db943","observation_id":"24787934-f6cc-4630-b1bd-907232ad4328","resolution":{"observed_at":"2026-08-06T15:18:11.656139Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T12:39:34.079317Z","title":"Safety layers in aligned large language models: The key to llm security","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.01414","last_updated":"2025-09-01T12:11:42Z","snapshot_observed_at":"2026-08-09T07:45:25.479566Z","submitted_at":"2025-09-01T12:11:42Z","title":"AttenTrack: Mobile User Attention Awareness Based on Context and External Distractions","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-05T12:39:34.079317Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2509.01414"},"observation_digest":"sha256:913ee378c6f099db2e89351ef430a787954c61844e514fc4c8917f09268aa8a9","observation_id":"584c895b-3c08-4de2-92b6-025324c04dff","resolution":{"observed_at":"2026-08-05T12:39:34.079317Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T06:00:51.098948Z","title":"Safety layers of aligned large language models: The key to llm security,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.04714","last_updated":"2025-09-05T00:02:17Z","snapshot_observed_at":"2026-08-09T01:25:34.681397Z","submitted_at":"2025-09-05T00:02:17Z","title":"ThumbnailTruth: A Multi-Modal LLM Approach for Detecting Misleading YouTube Thumbnails Across Diverse Cultural Settings","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-05T06:00:51.098948Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2509.04714"},"observation_digest":"sha256:d38931323bc1570e3492aaea99f579862c4b06936fb84298e19475e7716d60ce","observation_id":"69d67a9a-514b-438a-90db-0a0fd96c33e1","resolution":{"observed_at":"2026-08-05T06:00:51.098948Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-04T23:10:21.170218Z","title":"Safety layers in aligned large language models: The key to llm security,","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2509.06795","last_updated":"2025-09-08T15:24:33Z","snapshot_observed_at":"2026-08-06T11:18:33.345942Z","submitted_at":"2025-09-08T15:24:33Z","title":"Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraint","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-04T23:10:21.170218Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2509.06795"},"observation_digest":"sha256:ec15b84467335f4ce96b2b66c824674252e646ff7d59625c78a1a484ae867892","observation_id":"e084c66c-68b0-4f02-874b-7e9a41c635bd","resolution":{"observed_at":"2026-08-04T23:10:21.170218Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2511.06516","last_updated":"2026-06-28T18:20:13Z","snapshot_observed_at":"2026-08-03T23:21:36.698338Z","submitted_at":"2025-11-09T19:58:24Z","title":"You Had One Job: Per-Task Quantization Using LLMs' Hidden Representations","version":3},"reference_index":36,"source":"arxiv_source","source_observed_at":"2026-05-21T18:46:04.926179Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2511.06516"},"observation_digest":"sha256:245716da1603a8f02bd890aeb261ed5678ca750d77fd621afc0a660614e9a0d0","observation_id":"3b9e438f-ad96-431b-a1f4-ed0d470b50f6","resolution":{"observed_at":"2026-05-21T18:50:30.239698Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-03T23:21:39.668021Z","title":"Safety layers in aligned large language models: The key to llm security","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2511.06516","last_updated":"2026-06-28T18:20:13Z","snapshot_observed_at":"2026-08-03T23:21:36.698338Z","submitted_at":"2025-11-09T19:58:24Z","title":"You Had One Job: Per-Task Quantization Using LLMs' Hidden Representations","version":4},"reference_index":36,"source":"arxiv_source","source_observed_at":"2026-08-03T23:21:39.668021Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2511.06516"},"observation_digest":"sha256:6481680b906347dfc24ce829a109296d43df802c1fccf5beab1c40c4a6732181","observation_id":"1c1476a3-099a-4a33-9b61-d5606d6db2b1","resolution":{"observed_at":"2026-08-03T23:21:39.668021Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2602.15853","last_updated":"2026-04-27T00:25:03Z","snapshot_observed_at":"2026-07-31T20:33:52.305319Z","submitted_at":"2026-01-24T03:58:45Z","title":"A Lightweight Explainable Guardrail for Prompt Safety","version":2},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-05-16T11:46:14.226584Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2602.15853"},"observation_digest":"sha256:39dabbf4ca6c02892178cc656345b0a2ee13c03786e94c4d74fc83a651c20254","observation_id":"3481c5ea-0f6b-4fd7-9030-6b18cc4775ad","resolution":{"observed_at":"2026-05-16T11:47:49.345374Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2604.06247","last_updated":"2026-04-06T16:29:05Z","snapshot_observed_at":"2026-07-06T22:54:48.916799Z","submitted_at":"2026-04-06T16:29:05Z","title":"SALLIE: Safeguarding Against Latent Language & Image Exploits","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-05-10T19:04:46.426969Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2604.06247"},"observation_digest":"sha256:7339f4415369bafa3975619213b334c996e721225270ed18fafe40a70e084220","observation_id":"f77c2ff1-48ca-4d32-affb-ba58a833a3d3","resolution":{"observed_at":"2026-05-10T23:30:51.234243Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2604.11663","last_updated":"2026-04-13T16:11:38Z","snapshot_observed_at":"2026-07-06T22:59:59.220594Z","submitted_at":"2026-04-13T16:11:38Z","title":"Why Do Large Language Models Generate Harmful Content?","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-05-10T15:31:13.545599Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2604.11663"},"observation_digest":"sha256:c2f6a0004c17d459065561157b756e7eb75d2499a610a6c618af38a4fa13848b","observation_id":"139f1aa5-5823-4014-8e25-866da150b6a3","resolution":{"observed_at":"2026-05-11T10:21:04.499967Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2604.12384","last_updated":"2026-04-14T07:17:55Z","snapshot_observed_at":"2026-07-06T23:00:37.144068Z","submitted_at":"2026-04-14T07:17:55Z","title":"Preventing Safety Drift in Large Language Models via Coupled Weight and Activation Constraints","version":1},"reference_index":25,"source":"arxiv_source","source_observed_at":"2026-05-10T16:04:25.851592Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2604.12384"},"observation_digest":"sha256:989f107b38ce458a72c8dc9c1dcc3f042259092899aa63b8d899d048bf5f6f63","observation_id":"1587fbe5-177a-4d03-94cf-e5ceb4ca1ca0","resolution":{"observed_at":"2026-05-11T09:21:01.769674Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2604.18519","last_updated":"2026-04-20T17:17:07Z","snapshot_observed_at":"2026-07-06T23:05:21.974248Z","submitted_at":"2026-04-20T17:17:07Z","title":"LLM Safety From Within: Detecting Harmful Content with Internal Representations","version":1},"reference_index":33,"source":"arxiv_source","source_observed_at":"2026-05-10T04:33:54.058475Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2604.18519"},"observation_digest":"sha256:0647f36dd66f168d3a91cc4fd68ffaf454a6a64e041bb6a11e1e4dcd6072002f","observation_id":"d09d643a-6e56-4864-a8ff-6dafdace5ff8","resolution":{"observed_at":"2026-05-10T12:20:23.070753Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2605.10998","last_updated":"2026-05-09T15:52:29Z","snapshot_observed_at":"2026-08-03T01:46:31.309604Z","submitted_at":"2026-05-09T15:52:29Z","title":"Few-Shot Truly Benign DPO Attack for Jailbreaking LLMs","version":1},"reference_index":30,"source":"arxiv_source","source_observed_at":"2026-05-13T07:06:46.387088Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2605.10998"},"observation_digest":"sha256:b77a4656554aa904fb678adf665e49fecf6f7535e28c894c3953b76304c6170f","observation_id":"754ebd1a-ce5c-4832-a16a-f49f72bf005a","resolution":{"observed_at":"2026-05-13T07:07:26.976711Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2605.14514","last_updated":"2026-05-14T07:58:47Z","snapshot_observed_at":"2026-08-10T04:23:43.463729Z","submitted_at":"2026-05-14T07:58:47Z","title":"Defenses at Odds: Measuring and Explaining Defense Conflicts in Large Language Models","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-05-15T01:43:22.777232Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2605.14514"},"observation_digest":"sha256:055bf71165cbbd9c5f9efdcd40ec8d37e3010f51a5c5aa2a485c5d9136c00103","observation_id":"9fb2d29e-5e4b-4ff3-944f-99559e102d70","resolution":{"observed_at":"2026-05-15T01:43:27.337237Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2606.07335","last_updated":"2026-06-05T14:49:26Z","snapshot_observed_at":"2026-07-06T23:47:00.425715Z","submitted_at":"2026-06-05T14:49:26Z","title":"Defending Jailbreak Attacks on Large Language Models via Manifold Trajectory Kinetics","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-06-27T21:55:48.561400Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2606.07335"},"observation_digest":"sha256:4bca463bbb4862199d1848600f6b3081d6852ce2e252e51cea4646bb8453209d","observation_id":"21cc8787-5edb-4d31-ac65-ae4b255935fd","resolution":{"observed_at":"2026-07-02T17:37:14.893132Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2606.15980","last_updated":"2026-06-18T23:56:29Z","snapshot_observed_at":"2026-07-06T23:52:37.922109Z","submitted_at":"2026-06-14T19:07:22Z","title":"Do Activation Monitors Survive Model Updates? Benchmarking, Predicting, and Repairing Activation-Monitor Staleness","version":2},"reference_index":55,"source":"arxiv_source","source_observed_at":"2026-06-27T03:24:24.714121Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2606.15980"},"observation_digest":"sha256:56c024c3f012a4e8062506cfe5b6be1b021740fa7fabd3a38d89a9d651a8e86a","observation_id":"b5a85a15-ae91-47a9-b963-c642c9be3b3e","resolution":{"observed_at":"2026-06-27T03:30:26.988520Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2606.28153","last_updated":"2026-06-29T15:00:34Z","snapshot_observed_at":"2026-07-07T00:02:18.989389Z","submitted_at":"2026-06-26T14:51:16Z","title":"Robust Harmful Features Under Jailbreak Attacks: Mechanistic Evidence from Attention Head Specialization in Large Language Models","version":1},"reference_index":86,"source":"arxiv_source","source_observed_at":"2026-06-29T03:35:34.594617Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2606.28153"},"observation_digest":"sha256:f4041fe8fd8ad97d8073cc8d9d7c2f63ae84a1d8ee943d752fe11bbdcf24aa6d","observation_id":"986a73b8-f404-4c7f-a866-6e054d114540","resolution":{"observed_at":"2026-07-01T17:35:51.328901Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":"2408.17003","doi":"10.48550/arxiv.2408.17003","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Safety layers of aligned large language models: The key to llm security","venue":"arXiv (Cornell University)","work_id":"4807e5d2-329e-4c63-bbe1-67f643b30f78","year":2024},"citing_paper":{"arxiv_id":"2606.28153","last_updated":"2026-06-29T15:00:34Z","snapshot_observed_at":"2026-07-07T00:02:18.989389Z","submitted_at":"2026-06-26T14:51:16Z","title":"Robust Harmful Features Under Jailbreak Attacks: Mechanistic Evidence from Attention Head Specialization in Large Language Models","version":2},"reference_index":38,"source":"arxiv_source","source_observed_at":"2026-06-30T09:32:59.824110Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2606.28153"},"observation_digest":"sha256:d327db23ba650a586645fc9f8e3018df7b3c8a8fed7bcf8064ff12ccd8122f3d","observation_id":"9a408c1a-a20c-44a1-9484-7a1639be7c37","resolution":{"observed_at":"2026-06-30T09:34:34.453497Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-10T06:31:04.303077+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-01T17:41:11.909026Z","title":"arXiv preprint arXiv:2408.17003 , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17575","last_updated":"2026-07-20T05:42:25Z","snapshot_observed_at":"2026-08-08T20:41:24.949618Z","submitted_at":"2026-07-20T05:42:25Z","title":"A Dual-Hypothesis Reasoning Framework for LLM Guardrails","version":1},"reference_index":15,"source":"arxiv_source","source_observed_at":"2026-08-01T17:41:11.909026Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2607.17575"},"observation_digest":"sha256:d3342fa77383228158e3e5d537ce42d05abf89e0e9501e353b526b26d58db49f","observation_id":"1bc1f87d-385a-4433-91f2-33b429b9c815","resolution":{"observed_at":"2026-08-01T17:41:11.909026Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-01T12:04:24.102817Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.19683","last_updated":"2026-07-22T02:33:56Z","snapshot_observed_at":"2026-08-07T20:47:58.956911Z","submitted_at":"2026-07-22T02:33:56Z","title":"GhostPrompt: Cross-Image Adversarial Prompt for Vision-Language Models","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-01T12:04:24.102817Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2607.19683"},"observation_digest":"sha256:89e8a0dcc3453c016e28f59b670e5ff05a478d6ff23429d90f18df00669abd16","observation_id":"f418fcab-282e-49ce-a78c-99e4c6b4dd68","resolution":{"observed_at":"2026-08-01T12:04:24.102817Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-01T13:10:37.406289Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.22716","last_updated":"2026-07-21T15:52:30Z","snapshot_observed_at":"2026-08-06T12:10:57.542923Z","submitted_at":"2026-07-21T15:52:30Z","title":"Visual Token Compression Enhances Robustness of MLLMs","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-01T13:10:37.406289Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2607.22716"},"observation_digest":"sha256:c712eab39c0a1e4cff9d805b07d66b809c7bb19d33aded342cdb1fb2fcc7310a","observation_id":"37fc86a9-f4b7-4be8-b4bf-417450e1c4e0","resolution":{"observed_at":"2026-08-01T13:10:37.406289Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.17003","snapshot_observed_at":"2026-08-08T10:13:51.677511Z","title":"Safety Layers of Aligned Large Language Models: The Key to LLM Security.arXiv preprint arXiv:2408.17003, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2608.05578","last_updated":"2026-08-06T03:57:38Z","snapshot_observed_at":"2026-08-09T23:11:06.361553Z","submitted_at":"2026-08-06T03:57:38Z","title":"Detecting Safety Training Modification in Language Models via Activation Analysis","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-08T10:13:51.677511Z"},"links":{"cited_paper":"/paper/2408.17003","citing_paper":"/paper/2608.05578"},"observation_digest":"sha256:5f2877af4a6f4707d84f7363bd19ab05ca08258580b82c4ada4bf2af57faaa36","observation_id":"dd234149-1813-4a52-9f04-abdb222f3955","resolution":{"observed_at":"2026-08-08T10:13:51.677511Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"links":{"evidence":"/evidence","html":"/paper/2408.17003/citation-record","integrity":"/paper/2408.17003/integrity","json":"/paper/2408.17003/citation-record.json","paper":"/paper/2408.17003"},"outbound":[],"paper":{"arxiv_id":"2408.17003","last_updated":"2025-04-07T07:23:33Z","latest_version":5,"primary_category":"cs.CR","snapshot_observed_at":"2026-08-07T07:23:22.071348Z","submitted_at":"2024-08-30T04:35:59Z","title":"Safety Layers in Aligned Large Language Models: The Key to LLM Security"},"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":"2026-08-10T06:31:04.303077+00:00","source":"crossref"},{"observed_at":"2026-08-10T06:30:57.382061+00:00","source":"retraction_watch"}],"thesis":"As of 10 August 2026, this Paper Citation Record lists 0 of 0 outbound references and 27 inbound Pith citation observations for arXiv:2408.17003."}