{"as_of":"2026-08-12T18:16:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:841609aefab7e5ad53b598323d49aa4023a8e19e64d24bba089a65f4b21e606c","coverage":[{"denominator":38,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":38,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-11T20:30:10.363139Z","state":"measured"},{"denominator":42,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":42,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-12T06:34:41.77262+00:00","state":"measured"},{"denominator":4,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":4,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-06T13:05:39.686821Z","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":1,"observed_at":"2026-08-05T02:28:24.338817Z","source":"arxiv_reference"}],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.05734","snapshot_observed_at":"2026-08-06T13:05:39.686821Z","title":"Privagent: Agentic-based red-teaming for llm privacy leakage","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2507.21206","last_updated":"2025-07-28T17:58:12Z","snapshot_observed_at":"2026-08-08T10:48:52.442522Z","submitted_at":"2025-07-28T17:58:12Z","title":"Agentic Web: Weaving the Next Web with AI Agents","version":1},"reference_index":158,"source":"arxiv_source","source_observed_at":"2026-08-06T13:05:39.686821Z"},"links":{"cited_paper":"/paper/2412.05734","citing_paper":"/paper/2507.21206"},"observation_digest":"sha256:d2e684abefbaa4c35fffdc10eb61377d9eb58010a276416957eb6c8ca43491bc","observation_id":"319aba6c-0f4f-446f-8e42-c3303d458ac3","resolution":{"observed_at":"2026-08-06T13:05:39.686821Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"cited_work":{"arxiv_id":"2412.05734","doi":"10.48550/arxiv.2412.05734","metadata_source":"arxiv_reference","pith_arxiv_id":"2412.05734","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Available: https://arxiv.org/abs/2412.05734","venue":"arXiv (Cornell University)","work_id":"55e415a4-5e20-4bad-ab5a-891e2e144fa0","year":2024},"citing_paper":{"arxiv_id":"2603.09002","last_updated":"2026-04-26T14:13:48Z","snapshot_observed_at":"2026-08-11T02:37:54.478569Z","submitted_at":"2026-03-09T22:46:27Z","title":"Security Considerations for Multi-agent Systems","version":2},"reference_index":288,"source":"pdf_text","source_observed_at":"2026-05-15T14:12:14.160789Z"},"links":{"cited_paper":"/paper/2412.05734","citing_paper":"/paper/2603.09002"},"observation_digest":"sha256:7aaf841fad9cf4b5dc390f96c07c458aa9404fd4b0deb542e69381bb0124207f","observation_id":"2fb97ed7-fdff-44f2-853d-4f91537c8427","resolution":{"observed_at":"2026-05-15T14:15:55.840818Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"cited_work":{"arxiv_id":"2412.05734","doi":"10.48550/arxiv.2412.05734","metadata_source":"arxiv_reference","pith_arxiv_id":"2412.05734","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Available: https://arxiv.org/abs/2412.05734","venue":"arXiv (Cornell University)","work_id":"55e415a4-5e20-4bad-ab5a-891e2e144fa0","year":2024},"citing_paper":{"arxiv_id":"2606.06914","last_updated":"2026-06-05T05:26:42Z","snapshot_observed_at":"2026-08-06T16:14:45.008616Z","submitted_at":"2026-06-05T05:26:42Z","title":"DPAgent-in-the-Middle: Agentic Defense and Repair Against AI-Groomed Deceptive Patterns","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-06-27T21:56:41.335924Z"},"links":{"cited_paper":"/paper/2412.05734","citing_paper":"/paper/2606.06914"},"observation_digest":"sha256:18eaa53da8ac2b37ee04e163338ab41c65fed5a400167770e86171c605468718","observation_id":"c8949915-0c0c-447f-a86d-3a6d2f4d61a8","resolution":{"observed_at":"2026-07-02T17:37:14.703082Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"cited_work":{"arxiv_id":"2412.05734","doi":"10.48550/arxiv.2412.05734","metadata_source":"arxiv_reference","pith_arxiv_id":"2412.05734","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Available: https://arxiv.org/abs/2412.05734","venue":"arXiv (Cornell University)","work_id":"55e415a4-5e20-4bad-ab5a-891e2e144fa0","year":2024},"citing_paper":{"arxiv_id":"2606.07539","last_updated":"2026-04-29T17:39:36Z","snapshot_observed_at":"2026-08-02T12:30:34.186010Z","submitted_at":"2026-04-29T17:39:36Z","title":"Prompt Governance? On Governing Technologies Governed by Natural Language","version":1},"reference_index":246,"source":"pdf_text","source_observed_at":"2026-07-01T08:17:10.481202Z"},"links":{"cited_paper":"/paper/2412.05734","citing_paper":"/paper/2606.07539"},"observation_digest":"sha256:31a9c3bb6db45f5b606a5e6206f5c253b6ae6be81cf253f77fad643b1f40920b","observation_id":"ed30d0a5-1312-4c9e-9d73-1649233fea8f","resolution":{"observed_at":"2026-07-01T08:25:32.891728Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2412.05734/citation-record","integrity":"/paper/2412.05734/integrity","json":"/paper/2412.05734/citation-record.json","paper":"/paper/2412.05734"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-11T20:30:11.020955Z","title":"system prompt","venue":null,"work_id":"d3f5246f-f815-4360-a98a-6dc6d93ef52b","year":2025},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.357369Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:715126432b9661d40200527661dc68d9b78951e3977d7eccc7d1cc113d8f9c06","observation_id":"d7e2387f-84cc-4b6d-9fcd-d6f9a7cd7a10","resolution":{"observed_at":"2026-08-11T20:30:11.025748Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.06363","last_updated":"2024-09-25T19:48:39Z","snapshot_observed_at":"2026-08-11T03:37:22.279261Z","submitted_at":"2024-02-09T12:15:51Z","title":"StruQ: Defending Against Prompt Injection with Structured Queries","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.06363","snapshot_observed_at":"2026-08-11T20:30:10.206114Z","title":"Struq: Defending against prompt injection with structured queries","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.206114Z"},"links":{"cited_paper":"/paper/2402.06363","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:5073e2f1559ed289e100685d9d8c0dc910f6ed3046a898b120ee090c9dec2c92","observation_id":"bb90235e-3224-464d-8b9a-86ee358d668a","resolution":{"observed_at":"2026-08-11T20:30:10.206114Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1810.04805","last_updated":"2019-05-24T20:37:26Z","snapshot_observed_at":"2026-07-30T09:12:38.100527Z","submitted_at":"2018-10-11T00:50:01Z","title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1810.04805","snapshot_observed_at":"2026-08-11T20:30:10.210890Z","title":"Bert: Pre-training of deep bidirectional transformers for language understand- ing","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.210890Z"},"links":{"cited_paper":"/paper/1810.04805","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:c28ebe36ff37d76bd25547c2cb48bc0f9ab3a2c1d87253f129183ddf5836bcc2","observation_id":"3d3c2378-e518-4fba-87d7-9e67533fab0e","resolution":{"observed_at":"2026-08-11T20:30:10.210890Z","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":"2026-08-11T20:30:10.222173Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.222173Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:c090f2845bd2594b7522ac318eee7ab11d215a3e6f4b01d5d099290c38ec2e57","observation_id":"70c6c925-aa4c-42bf-bf63-adf9fb5b0654","resolution":{"observed_at":"2026-08-11T20:30:10.222173Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1708.06733","last_updated":"2019-03-11T20:45:33Z","snapshot_observed_at":"2026-08-12T15:43:59.037380Z","submitted_at":"2017-08-22T17:31:54Z","title":"BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1708.06733","snapshot_observed_at":"2026-08-11T20:30:10.226536Z","title":"Tianyu Gu, Brendan Dolan-Gavitt, and Siddharth Garg","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.226536Z"},"links":{"cited_paper":"/paper/1708.06733","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:321a500c057e4ecaaf9bc9ac64284235028d798652b0b2fa686873f7062c6d9c","observation_id":"e2205c4c-40c8-447a-b569-9040e9812273","resolution":{"observed_at":"2026-08-11T20:30:10.226536Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2312.06674","last_updated":"2023-12-07T19:40:50Z","snapshot_observed_at":"2026-07-06T17:00:00.321552Z","submitted_at":"2023-12-07T19:40:50Z","title":"Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2312.06674","snapshot_observed_at":"2026-08-11T20:30:10.235857Z","title":"Llama guard: Llm- based input-output safeguard for human-ai conversations","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.235857Z"},"links":{"cited_paper":"/paper/2312.06674","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:80b5e58b782c80605ae69072e13e88ca8fc3c9887cf9cf9a3af21fbe2e1c4f97","observation_id":"6c71d5aa-d421-46f1-b89c-9f29c873d726","resolution":{"observed_at":"2026-08-11T20:30:10.235857Z","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":"2026-08-11T20:30:10.240270Z","title":"11 Published as a conference paper at COLM 2025 Kevin Liu [@kliu128]","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.240270Z"},"links":{"cited_paper":"/paper/2310.06825","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:274b7b835474b7f9934ba1fdf30d7dc083a603b8dada03ede6d0109febca5974","observation_id":"1ede0a9b-6611-4de4-bede-d538a7a016bb","resolution":{"observed_at":"2026-08-11T20:30:10.240270Z","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":"2026-08-04T19:12:56.970085Z","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":"2026-08-11T20:30:10.244470Z","title":"Deepincep- tion: Hypnotize large language model to be jailbreaker","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.244470Z"},"links":{"cited_paper":"/paper/2311.03191","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:0a99f067d8b8f81c41f3d1bb78bf7eb1f8f23f85a59786c78717103e15757189","observation_id":"2bfaad37-7870-47a8-91c7-a86c2a4ddf27","resolution":{"observed_at":"2026-08-11T20:30:10.244470Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.04451","last_updated":"2024-03-20T21:34:56Z","snapshot_observed_at":"2026-08-06T02:57:30.438059Z","submitted_at":"2023-10-03T19:44:37Z","title":"AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.04451","snapshot_observed_at":"2026-08-11T20:30:10.253096Z","title":"Autodan: Generating stealthy jailbreak prompts on aligned large language models","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.253096Z"},"links":{"cited_paper":"/paper/2310.04451","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:7470e9ccbbcee810f00b04f63501163d63eba9354c9eb37a7e810eafb5e815cf","observation_id":"2a6aae8e-936f-4ba7-ac36-c3a48f1a5e14","resolution":{"observed_at":"2026-08-11T20:30:10.253096Z","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":"2026-08-11T20:30:11.086949Z","title":"Meta-llama: Prompt guard, 2024c","venue":null,"work_id":"e24a5e06-8ef0-4cec-ae77-88b8c41e67e8","year":2024},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.256790Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:b9763ab31120d5481670b4262cf091209f7a35cd341908b6915234beeb9a2267","observation_id":"be125807-3ea2-4675-80f1-1a8bfb197053","resolution":{"observed_at":"2026-08-11T20:30:11.090591Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2405.16783","last_updated":"2024-05-27T03:10:57Z","snapshot_observed_at":"2026-08-11T02:16:38.672171Z","submitted_at":"2024-05-27T03:10:57Z","title":"TrojFM: Resource-efficient Backdoor Attacks against Very Large Foundation Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2405.16783","snapshot_observed_at":"2026-08-11T20:30:10.264677Z","title":"Trojfm: Resource-efficient backdoor attacks against very large foundation models","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.264677Z"},"links":{"cited_paper":"/paper/2405.16783","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:a46d30243ce24884ad32ac09e37ea704c1adbc473054bc330b8b2f7cbebc6fba","observation_id":"fb1744fc-47f6-4554-a887-855d2474867a","resolution":{"observed_at":"2026-08-11T20:30:10.264677Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.02309","last_updated":"2024-02-04T01:29:24Z","snapshot_observed_at":"2026-08-07T07:47:20.838090Z","submitted_at":"2024-02-04T01:29:24Z","title":"Jailbreaking Attack against Multimodal Large Language Model","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.02309","snapshot_observed_at":"2026-08-11T20:30:10.268953Z","title":"Jailbreaking attack against multimodal large language model","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.268953Z"},"links":{"cited_paper":"/paper/2402.02309","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:5f6037d3ab2cb486414128bb03d731caf29e8ea0652cfa33e8072e69dfe4fbdd","observation_id":"f30726b4-e836-4375-93e4-62e7e351d7c2","resolution":{"observed_at":"2026-08-11T20:30:10.268953Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.17623","last_updated":"2023-11-24T01:45:16Z","snapshot_observed_at":"2026-07-06T16:39:08.850847Z","submitted_at":"2023-10-26T17:43:13Z","title":"Proving Test Set Contamination in Black Box Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.17623","snapshot_observed_at":"2026-08-11T20:30:10.273799Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.273799Z"},"links":{"cited_paper":"/paper/2310.17623","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:4a48cf7fbb1e105eef30d09c0e385f0ed925ae110577376f0428d11c2b973f0b","observation_id":"09bf5eeb-7e7e-414f-a8c6-e6df7b607cf0","resolution":{"observed_at":"2026-08-11T20:30:10.273799Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2211.09527","last_updated":"2022-11-17T13:43:20Z","snapshot_observed_at":"2026-07-06T14:19:47.424778Z","submitted_at":"2022-11-17T13:43:20Z","title":"Ignore Previous Prompt: Attack Techniques For Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2211.09527","snapshot_observed_at":"2026-08-11T20:30:10.283794Z","title":"Ignore previous prompt: Attack techniques for language models","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.283794Z"},"links":{"cited_paper":"/paper/2211.09527","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:3174a8f2849554286bab2446c6fa2428c9aa266f38dd26034ce88369e1ddfbd7","observation_id":"2afaa76d-82d9-4fcb-b71f-feeecd0c52ff","resolution":{"observed_at":"2026-08-11T20:30:10.283794Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1707.06347","last_updated":"2017-08-28T09:20:06Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2017-07-20T02:32:33Z","title":"Proximal Policy Optimization Algorithms","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1707.06347","snapshot_observed_at":"2026-08-11T20:30:10.291879Z","title":"John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.291879Z"},"links":{"cited_paper":"/paper/1707.06347","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:e5c2c3485a9e5570370c0ad3867e76f5066c7a0847da4dcbfb5d09a1e7446ea6","observation_id":"d7550f68-7d3f-4a78-8c77-64aa244de3f0","resolution":{"observed_at":"2026-08-11T20:30:10.291879Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2304.12298","last_updated":"2023-02-21T08:59:22Z","snapshot_observed_at":"2026-07-06T15:19:28.033750Z","submitted_at":"2023-02-21T08:59:22Z","title":"BadGPT: Exploring Security Vulnerabilities of ChatGPT via Backdoor Attacks to InstructGPT","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2304.12298","snapshot_observed_at":"2026-08-11T20:30:10.297459Z","title":"Badgpt: Exploring security vulnerabilities of chatgpt via backdoor attacks to instructgpt","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.297459Z"},"links":{"cited_paper":"/paper/2304.12298","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:11ccc880e3aaff3f8da9c41fb6661d004a4dfeeac57aa4127e4f6f131feb8792","observation_id":"88360933-99a7-4541-be0a-1230fb97414d","resolution":{"observed_at":"2026-08-11T20:30:10.297459Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2406.12814","last_updated":"2025-02-04T20:02:17Z","snapshot_observed_at":"2026-08-08T10:00:10.515051Z","submitted_at":"2024-06-18T17:32:48Z","title":"Dissecting Adversarial Robustness of Multimodal LM Agents","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2406.12814","snapshot_observed_at":"2026-08-11T20:30:10.308201Z","title":"Adversarial attacks on multimodal agents","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.308201Z"},"links":{"cited_paper":"/paper/2406.12814","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:71dfa7862a963b321036cd3afeaad8f390aea4fce904dd809186c916d6d9a449","observation_id":"d87bf9cb-fe61-484b-bfe6-8b313980a183","resolution":{"observed_at":"2026-08-11T20:30:10.308201Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.19200","last_updated":"2025-06-12T13:56:26Z","snapshot_observed_at":"2026-08-05T03:23:44.922485Z","submitted_at":"2024-02-29T14:30:28Z","title":"PRSA: Prompt Stealing Attacks against Real-World Prompt Services","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.19200","snapshot_observed_at":"2026-08-11T20:30:10.312732Z","title":"Prsa: Prompt reverse stealing attacks against large language models","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.312732Z"},"links":{"cited_paper":"/paper/2402.19200","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:d0e435ee4f6457f70e389ba91256c0de48d05392a118cc159f635616dfe75485","observation_id":"33e5bc8a-6939-4ef2-8915-555354939987","resolution":{"observed_at":"2026-08-11T20:30:10.312732Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2210.03629","last_updated":"2023-03-10T01:00:17Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2022-10-06T01:00:32Z","title":"ReAct: Synergizing Reasoning and Acting in Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2210.03629","snapshot_observed_at":"2026-08-11T20:30:10.317006Z","title":"React: Synergizing reasoning and acting in language models","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.317006Z"},"links":{"cited_paper":"/paper/2210.03629","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:2c4487b799e88cc3a0c7658790801fec233cbfbd1ec5d52c41a840d2665d031c","observation_id":"803d00f7-fe4c-46ae-8752-8fbcab1d4df0","resolution":{"observed_at":"2026-08-11T20:30:10.317006Z","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":"2026-08-11T20:30:11.060435Z","title":"Privacy risk in machine learning: Analyzing the connection to overfitting","venue":null,"work_id":"51d118e6-3c73-47fe-a0d9-da0e1f04dee5","year":2018},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.321400Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:d914e39a8044f55434ebb885a2962bf1f18be91edec9cb234d9645f774236314","observation_id":"683bf443-ef1d-4bee-bc39-ba2b26cfaa0e","resolution":{"observed_at":"2026-08-11T20:30:11.064605Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2309.10253","last_updated":"2024-06-27T16:01:27Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-09-19T02:19:48Z","title":"GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2309.10253","snapshot_observed_at":"2026-08-11T20:30:10.325968Z","title":"Gptfuzzer: Red teaming large language models with auto-generated jailbreak prompts","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.325968Z"},"links":{"cited_paper":"/paper/2309.10253","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:01c78e2e7acab7fbe6dad4f701779ce075940b765c01ac3eaa5754b3f4feafc8","observation_id":"137707ef-536c-4e79-a23a-8eeff5f82eae","resolution":{"observed_at":"2026-08-11T20:30:10.325968Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2405.20653","last_updated":"2025-06-17T03:03:45Z","snapshot_observed_at":"2026-08-06T16:41:41.743961Z","submitted_at":"2024-05-31T07:41:03Z","title":"Mind the Inconspicuous: Revealing the Hidden Weakness in Aligned LLMs' Refusal Boundaries","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2405.20653","snapshot_observed_at":"2026-08-11T20:30:10.330931Z","title":"Enhancing jailbreak attack against large language models through silent tokens","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.330931Z"},"links":{"cited_paper":"/paper/2405.20653","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:f3825139bd3a4cd898f7b4b02687446884948fbeff8a4c18f44f733930a422b9","observation_id":"3704a8bc-9454-447f-929a-3d2959eb0187","resolution":{"observed_at":"2026-08-11T20:30:10.330931Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2307.06865","last_updated":"2024-08-07T22:05:01Z","snapshot_observed_at":"2026-08-07T22:01:02.608714Z","submitted_at":"2023-07-13T16:15:08Z","title":"Effective Prompt Extraction from Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2307.06865","snapshot_observed_at":"2026-08-11T20:30:10.340189Z","title":"Prompts should not be seen as secrets: Systematically measuring prompt extraction attack success","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.340189Z"},"links":{"cited_paper":"/paper/2307.06865","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:054b64dd8b2ff3caadbbd06b7a294816645f7d12d6dc2bc48166c8fea2bc87d0","observation_id":"626a6b9a-8208-42e1-b331-411401bdd29d","resolution":{"observed_at":"2026-08-11T20:30:10.340189Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2307.15043","last_updated":"2023-12-20T20:48:57Z","snapshot_observed_at":"2026-08-12T09:06:50.363435Z","submitted_at":"2023-07-27T17:49:12Z","title":"Universal and Transferable Adversarial Attacks on Aligned Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2307.15043","snapshot_observed_at":"2026-08-11T20:30:10.344422Z","title":"Universal and transferable adversarial attacks on aligned language models","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.344422Z"},"links":{"cited_paper":"/paper/2307.15043","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:be920e1bff7469e228a71d4865771cd3dc8b38ffab5cccc636433933472d0dfc","observation_id":"a0cfef43-7b89-42bc-a420-745325676ed2","resolution":{"observed_at":"2026-08-11T20:30:10.344422Z","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":"2026-08-11T20:30:11.047601Z","title":"Please generate a prompt for me","venue":null,"work_id":"ec569b79-a0d7-4947-9893-8c3f29f5f61e","year":2025},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.348516Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:628f68eecf843d09cd4f2aa85a8c5a2eb066e3c91b0ab81b4f3c3ea47d269aad","observation_id":"214306e9-8c58-4f57-a934-1195dccc8d96","resolution":{"observed_at":"2026-08-11T20:30:11.051937Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+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":"2026-08-11T20:30:11.034377Z","title":"Where is the prompt from? (e.g","venue":null,"work_id":"de8ac1b4-68ba-40d2-b8b2-5bbef07fb401","year":2025},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.353562Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:be99eab27dbadcfc7ece1d366e16c54fec0aa12f0e4071b0fabb5ce072f0568a","observation_id":"fbba056c-a31d-4ebb-95cc-9f338d7d0eb5","resolution":{"observed_at":"2026-08-11T20:30:11.038633Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+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":"2026-08-11T20:30:11.004681Z","title":"Both AWS Lambda and the functions running on the service provide predictable and reliable operational performance","venue":null,"work_id":"db1d3986-fbfd-4c17-94c8-158e7843c566","year":2025},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.363139Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:3ea131969593825e74c9dc834bd7d547c0f9d66f3bfa8340b3ce566328cc3a63","observation_id":"ce7f61cd-d3cc-458d-92ad-0fb45ebb2347","resolution":{"observed_at":"2026-08-11T20:30:11.011712Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2311.17035","last_updated":"2023-11-28T18:47:03Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-11-28T18:47:03Z","title":"Scalable Extraction of Training Data from (Production) Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2311.17035","snapshot_observed_at":"2026-08-11T20:30:10.260939Z","title":"Scalable extraction of training data from (production) language models","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":1990,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.260939Z"},"links":{"cited_paper":"/paper/2311.17035","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:d8e9909297c8419147ecb28448cbfc8759f1308aa1f7d638772dcdd670327d3b","observation_id":"cc451ab1-86f5-43c2-89ee-a7843157067c","resolution":{"observed_at":"2026-08-11T20:30:10.260939Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2404.16873","last_updated":"2025-06-02T18:59:01Z","snapshot_observed_at":"2026-08-05T10:26:02.067012Z","submitted_at":"2024-04-21T22:18:13Z","title":"AdvPrompter: Fast Adaptive Adversarial Prompting for LLMs","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2404.16873","snapshot_observed_at":"2026-08-11T20:30:10.279162Z","title":"Advprompter: Fast adaptive adversarial prompting for llms","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2002,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.279162Z"},"links":{"cited_paper":"/paper/2404.16873","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:402677396aa0fc3a2c09b996d6511a04daf19ec171710da6ac368ef56889d91f","observation_id":"48e175c0-10d4-43b7-8d8a-18670780c9f9","resolution":{"observed_at":"2026-08-11T20:30:10.279162Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.07403","last_updated":"2024-07-12T03:58:05Z","snapshot_observed_at":"2026-08-10T04:33:40.668262Z","submitted_at":"2024-07-10T06:57:58Z","title":"A Survey of Attacks on Large Vision-Language Models: Resources, Advances, and Future Trends","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.07403","snapshot_observed_at":"2026-08-11T20:30:10.249344Z","title":"A survey of attacks on large vision-language models: Resources, advances, and future trends","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2004,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.249344Z"},"links":{"cited_paper":"/paper/2407.07403","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:ab26aece71646cceb924ff7758eaa123ead93e178e4da24d717aa05bcbe5b6dc","observation_id":"a1df7941-06a3-4455-aea7-e5fb397b53fc","resolution":{"observed_at":"2026-08-11T20:30:10.249344Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2403.02691","last_updated":"2024-08-04T04:52:35Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-03-05T06:21:45Z","title":"InjecAgent: Benchmarking Indirect Prompt Injections in Tool-Integrated Large Language Model Agents","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.02691","snapshot_observed_at":"2026-08-11T20:30:10.335485Z","title":"Injecagent: Benchmarking indirect prompt injections in tool-integrated large language model agents","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2007,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.335485Z"},"links":{"cited_paper":"/paper/2403.02691","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:6f5ceb552842d12ff4638882dc748a38073b521a937b50cac4271c5396f636ff","observation_id":"6f82f768-c3f9-44ee-9515-76f0b12691e1","resolution":{"observed_at":"2026-08-11T20:30:10.335485Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2405.06823","last_updated":"2025-06-13T02:24:21Z","snapshot_observed_at":"2026-08-11T10:00:27.407495Z","submitted_at":"2024-05-10T21:52:34Z","title":"PLeak: Prompt Leaking Attacks against Large Language Model Applications","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2405.06823","snapshot_observed_at":"2026-08-11T20:30:10.231146Z","title":"Pleak: Prompt leaking attacks against large language model applications","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2017,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.231146Z"},"links":{"cited_paper":"/paper/2405.06823","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:d34aad25001ec37b1b1b5b4e37488b32ef36f590699bb2f38fc2c2d92084dbd9","observation_id":"4f1e3c20-0a53-4756-b839-5fe9e336e83b","resolution":{"observed_at":"2026-08-11T20:30:10.231146Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.09910","last_updated":"2024-06-25T10:33:41Z","snapshot_observed_at":"2026-08-11T03:10:30.141160Z","submitted_at":"2024-02-15T12:17:15Z","title":"DE-COP: Detecting Copyrighted Content in Language Models Training Data","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.09910","snapshot_observed_at":"2026-08-11T20:30:10.217330Z","title":"De-cop: Detecting copyrighted content in language models training data","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2018,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.217330Z"},"links":{"cited_paper":"/paper/2402.09910","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:d9a3fc45559bd544a15eb4645eaa3ecdebd414cee3c15a1f9b725867c8760d36","observation_id":"65af906d-2712-410d-b18a-1c25d15ad5de","resolution":{"observed_at":"2026-08-11T20:30:10.217330Z","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":"2026-08-11T20:30:10.303552Z","title":"Trustllm: Trustworthiness in large language models","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2019,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.303552Z"},"links":{"cited_paper":"/paper/2401.05561","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:149112549a6c284a0214857dca01da71d5a764d0bb4f7c1fffa59884fdf6a14b","observation_id":"afc50eb7-1c38-486b-804b-04a94de65787","resolution":{"observed_at":"2026-08-11T20:30:10.303552Z","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":"2026-08-11T20:30:10.194907Z","title":"Membership inference attacks from first principles","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2021,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.194907Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:4ea939916209763a121e418e4833fdd504d2672f0d981af9dcb6d38dcf8f5a2c","observation_id":"f5c21190-52e4-4882-8256-c576a93bac38","resolution":{"observed_at":"2026-08-11T20:30:10.194907Z","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":"2026-08-11T20:30:11.073531Z","title":null,"venue":null,"work_id":"3356e001-c293-4733-8056-b119dfa91a66","year":2025},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2022,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.287980Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:54a70c18da49f470b15d195743dcf41b32b5a7f265db3634fa6032a171f80170","observation_id":"e309b414-39c0-48f2-9532-2905ff2dc40b","resolution":{"observed_at":"2026-08-11T20:30:11.077379Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+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":"2026-08-11T20:30:11.110271Z","title":"claude model","venue":null,"work_id":"891518ca-f5a7-4037-a8af-14299a8c573f","year":2022},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2023,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.189606Z"},"links":{"citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:1d2d218376a71f007e7bbbb1605af660c590b2e664586e902f1d8a9c5b8dc266","observation_id":"3f79c285-14c8-46a3-89c9-01a46a979458","resolution":{"observed_at":"2026-08-11T20:30:11.115557Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-12T06:34:41.77262+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2403.06634","last_updated":"2024-07-09T17:44:00Z","snapshot_observed_at":"2026-08-10T12:12:16.686272Z","submitted_at":"2024-03-11T11:46:12Z","title":"Stealing Part of a Production Language Model","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.06634","snapshot_observed_at":"2026-08-11T20:30:10.200969Z","title":"Patrick Chao, Alexander Robey, Edgar Dobriban, Hamed Hassani, George J Pappas, and Eric Wong","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage","version":2},"reference_index":2024,"source":"pdf_text","source_observed_at":"2026-08-11T20:30:10.200969Z"},"links":{"cited_paper":"/paper/2403.06634","citing_paper":"/paper/2412.05734"},"observation_digest":"sha256:6eab2b4ad48a235fb2baa9821ee864b8c486c056265767e91a6e5652977bab43","observation_id":"1f9a7a37-d4ad-43c8-a9e1-3f578c733ef6","resolution":{"observed_at":"2026-08-11T20:30:10.200969Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2412.05734","last_updated":"2025-08-08T09:27:21Z","latest_version":2,"primary_category":"cs.CR","snapshot_observed_at":"2026-08-12T08:07:50.886221Z","submitted_at":"2024-12-07T20:09:01Z","title":"LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage"},"reference_resolution":{"displayed":38,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":31,"verified_exact":0,"verified_fuzzy":7},"total_outbound_references":38},"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":"2026-08-12T06:34:41.77262+00:00","source":"crossref"},{"observed_at":"2026-08-12T06:34:36.333875+00:00","source":"retraction_watch"}],"thesis":"As of 12 August 2026, this Paper Citation Record lists 38 of 38 outbound references and 4 inbound Pith citation observations for arXiv:2412.05734."}