{"as_of":"2026-08-03T20:31:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:a99d63b0a067e05b5a74967ea86c47b51ff92ac51e0c20c4b82e7bc4ef80bfe6","coverage":[{"denominator":65,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":65,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-02T06:50:15.020939Z","state":"measured"},{"denominator":65,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":65,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-03T06:30:56.289259+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2607.13088/citation-record","integrity":"/paper/2607.13088/integrity","json":"/paper/2607.13088/citation-record.json","paper":"/paper/2607.13088"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-02T06:50:13.010889Z","title":"Security and privacy challenges of large language models: A survey.ACM Computing Surveys, 57(6):1–39, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:13.010889Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:135520b8df9905b81ddf0af8178e21da1a6f8f04042a39813d7bb1ae0d5068be","observation_id":"a922672d-5224-4d88-a345-6eddc15155b8","resolution":{"observed_at":"2026-08-02T06:50:13.010889Z","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":"2026-08-02T06:50:13.102954Z","title":null,"venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:13.102954Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:bf2fa3a57181f05ebb5d58c4dc213c7848b3cd018928d7aac8f6709b426fd3b5","observation_id":"6d577685-e90a-41a6-8ae1-cdb9881a11d6","resolution":{"observed_at":"2026-08-02T06:50:13.102954Z","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":"2026-08-02T06:50:13.205852Z","title":"Department of Health and Human Services","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:13.205852Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:ba9324dd49f466be2f624a68be7849533720d391225245df4db93706d2d27bcc","observation_id":"2a517f81-26a5-4fcd-b2d7-249a8a58d79d","resolution":{"observed_at":"2026-08-02T06:50:13.205852Z","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":"2026-08-02T06:50:13.491779Z","title":"Regulation (EU) 2024/1689 of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act)","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:13.491779Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:1f86003cb2362dcde2627da73148188aedfce5c066c6cf57c85abdc15b4f3392","observation_id":"1b2cbe34-ab2f-4136-a24d-8af7345cc739","resolution":{"observed_at":"2026-08-02T06:50:13.491779Z","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":"2026-08-02T06:50:13.647492Z","title":"Removing Barriers to American Leadership in Artificial Intelligence","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:13.647492Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:0cafd95fa12a2b85b18e95cd36e6e4ce60c123998d37a8177bf97cc89c0235e3","observation_id":"fb9a3507-15a3-4772-b0b5-25a4c9f61d73","resolution":{"observed_at":"2026-08-02T06:50:13.647492Z","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":"2026-08-02T06:50:13.769394Z","title":"Strengthening our frontier safety framework","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:13.769394Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:9f9bb846b1404b2bfa8cc32f8d0ffd2a35eeb2030a0510bd467a2b25224b4a9f","observation_id":"21381d3b-51a1-46d8-ab7b-86d29a4870d3","resolution":{"observed_at":"2026-08-02T06:50:13.769394Z","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":"2026-08-02T06:50:13.969226Z","title":"Deploying llm transformer on edge computing devices: A survey of strategies, challenges, and future directions.AI, 7(1):15, 2026","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:13.969226Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:212bdb404617b595f4524ca4fd3b551bd17364b1c2b801455cf580dc8018f12d","observation_id":"d8cd93f3-4f1a-412a-b31a-5c5f762b48c5","resolution":{"observed_at":"2026-08-02T06:50:13.969226Z","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":"2026-08-02T06:50:14.081695Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.081695Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:7495e7203e7a30124aba2032abb27536be149cbc1baee7279dfabec77fc7102d","observation_id":"f93c1a17-2824-4ed0-9be1-ec68c7debd82","resolution":{"observed_at":"2026-08-02T06:50:14.081695Z","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":"2026-08-02T06:50:14.198426Z","title":"Efficient memory management for large language model serving with pagedattention","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.198426Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:cfcfb9519c1557a5b6b245f72f0b9c1f5331b8137b68c05a7770b09b622a0aa3","observation_id":"3cd713f8-830e-4204-8099-193607921a5c","resolution":{"observed_at":"2026-08-02T06:50:14.198426Z","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":"2026-08-02T06:50:14.303849Z","title":"Advancing practical homomorphic encryption for federated learning: Theoretical guarantees and efficiency optimizations.arXiv preprint arXiv:2509.20476, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.303849Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:3c3d18990a6cd01861ccc27606452d17dd04f876d5dfdbd75a9f366e6422980a","observation_id":"76499889-b987-4f30-93af-051bc233843c","resolution":{"observed_at":"2026-08-02T06:50:14.303849Z","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":"2026-08-02T06:50:14.442437Z","title":"Zeroquant: Efficient and affordable post- training quantization for large-scale transformers.Advances in neural information processing systems, 35:27168–27183, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.442437Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:68c7d9b5a66b5b3e9c04345bdee5ebb3510de3020dd00276488a8095e1ed479a","observation_id":"54b4d34a-eff4-499c-8542-12532d7b1d55","resolution":{"observed_at":"2026-08-02T06:50:14.442437Z","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":"2026-08-02T06:50:14.601915Z","title":"Exploiting llm quantization.Advances in Neural Information Processing Systems, 37:41709–41732, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.601915Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:fb8216d8ac3efa3096f115d71a868733e47771e7fbc49dae5df06a1a97f87bd8","observation_id":"f74e88df-bd47-4335-b08e-607fa27abc0a","resolution":{"observed_at":"2026-08-02T06:50:14.601915Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1412.6572","last_updated":"2015-03-20T20:19:16Z","snapshot_observed_at":"2026-07-06T04:04:16.777653Z","submitted_at":"2014-12-20T01:17:12Z","title":"Explaining and Harnessing Adversarial Examples","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1412.6572","snapshot_observed_at":"2026-08-02T06:50:14.781159Z","title":"Explaining and harnessing adversarial examples.arXiv preprint arXiv:1412.6572, 2014","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.781159Z"},"links":{"cited_paper":"/paper/1412.6572","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:89276a8e7b64750acc5906d0c684d1dc3125ef685b2df5c62fa3e29a36e2d4c2","observation_id":"8bef5eab-2f20-4152-a455-477850d34f6b","resolution":{"observed_at":"2026-08-02T06:50:14.781159Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1711.00449","last_updated":"2018-01-31T17:03:31Z","snapshot_observed_at":"2026-07-06T06:07:16.807271Z","submitted_at":"2017-11-01T17:28:26Z","title":"Attacking Binarized Neural Networks","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1711.00449","snapshot_observed_at":"2026-08-02T06:50:14.896441Z","title":"Attacking binarized neural networks.arXiv preprint arXiv:1711.00449, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.896441Z"},"links":{"cited_paper":"/paper/1711.00449","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:a37759f30f3ba1d0ed6856383a86fad02b87282d6b2b065b47abb6e1a52b2414","observation_id":"385e11d0-06ce-46f2-9aa3-b6deae1f5afb","resolution":{"observed_at":"2026-08-02T06:50:14.896441Z","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":"2026-08-02T06:50:14.899199Z","title":"Synthesizing robust adversarial examples","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.899199Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:9549434bb2200762da9e52fd3d7f0fd2bf1a572219d5720fee98a0c42dc23e02","observation_id":"1ece4f88-45ae-4db5-b916-eb355ecf00b6","resolution":{"observed_at":"2026-08-02T06:50:14.899199Z","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":"2026-08-02T06:50:14.902096Z","title":"Quantization aware attack: Enhancing transferable adversarial attacks by model quantization","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.902096Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:319e7a5356501b787eae61574898e2e5d630dfb53d4598ab33b3a7dfee54a8dc","observation_id":"8f1dbe98-401c-4ad7-a6ed-0ee53e0b7228","resolution":{"observed_at":"2026-08-02T06:50:14.902096Z","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":"2026-08-02T06:50:14.904935Z","title":"On jailbreaking quantized language models through fault injection attacks","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.904935Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:85db168b9eedd84dbf3abd20ceb73b8ecd12013ec9f5875e4f1ae386b574da67","observation_id":"f9707a9e-7d3e-4876-9480-ba6fe09395d7","resolution":{"observed_at":"2026-08-02T06:50:14.904935Z","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":"2026-08-02T06:50:14.907375Z","title":"Quantization-based jailbreaking vulnerability analysis: A study on performance and safety of the llama3-8b-instruct model.IEEE Access, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.907375Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:d3214f4796b2f8e26f2cc77d29e7acf4c2d4eab4ef6b7691a21855651fca6a2c","observation_id":"98c800fa-02e5-44c4-ab27-439967470c2c","resolution":{"observed_at":"2026-08-02T06:50:14.907375Z","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":"2026-08-02T06:50:14.909827Z","title":"Flipping bits in memory without accessing them: An experimental study of dram disturbance errors.ACM SIGARCH Computer Architecture News, 42(3):361–372, 2014","venue":null,"work_id":null,"year":2014},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.909827Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:dbe19c5d6e8743add96ec398f863090b14003b6da29ff8c6a51217c56a18cefc","observation_id":"c786fd67-6811-44e5-a0e4-47fdd0fc3a66","resolution":{"observed_at":"2026-08-02T06:50:14.909827Z","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":"2026-08-02T06:50:14.912258Z","title":"Bit-flip attack: Crushing neural network with progressive bit search","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.912258Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:9dd4cf66dfbd5720a375a5a6836619fdd32ed4d7e4770c5438bed09b93fe9ccc","observation_id":"b07c5533-8b8a-40a5-946c-de087c1f61ab","resolution":{"observed_at":"2026-08-02T06:50:14.912258Z","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":"2026-08-02T06:50:14.915099Z","title":"Deep reinforcement learning from human prefer- ences.Advances in neural information processing systems, 30, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.915099Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:c13a2c81fee9e06cce89b215d0406de7707790f3ba98d85c02121d017949cdcc","observation_id":"b0104b93-7d1f-4b2f-abe6-8026f745c0bf","resolution":{"observed_at":"2026-08-02T06:50:14.915099Z","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":"2026-08-02T06:50:14.917641Z","title":"Learning to summarize with human feedback.Advances in neural information processing systems, 33:3008–3021, 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.917641Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:7302c9a082d0d4d5e22dfa249ab14a1b71b7dd4c77df40450a6cdd9787aa9ea4","observation_id":"95844534-4e5a-471e-9851-714a65f47d10","resolution":{"observed_at":"2026-08-02T06:50:14.917641Z","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":"2026-08-02T06:50:14.920032Z","title":"Direct preference optimization: Your language model is secretly a reward model.Advances in neural information processing systems, 36:53728–53741, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.920032Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:1305aa27fee6738db8daef6ba2ecdd13e2ea537a1abf0f31748f1b038a7dc7f3","observation_id":"6036cab6-df6d-4bc5-acdd-f3e0541b911f","resolution":{"observed_at":"2026-08-02T06:50:14.920032Z","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":"2026-08-02T06:50:14.922432Z","title":"Safety alignment should be made more than just a few tokens deep","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.922432Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:3d165b6886444587f20104aeac0232f87c9478942ad1cdac9ba907abab068ff0","observation_id":"77e60fdf-6646-47d2-aaeb-7d4a8650fda7","resolution":{"observed_at":"2026-08-02T06:50:14.922432Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1803.03635","last_updated":"2019-03-04T15:51:11Z","snapshot_observed_at":"2026-07-06T06:27:32.857286Z","submitted_at":"2018-03-09T18:51:28Z","title":"The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1803.03635","snapshot_observed_at":"2026-08-02T06:50:14.924793Z","title":"The lottery ticket hypoth- esis: Finding sparse, trainable neural networks.arXiv preprint arXiv:1803.03635, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.924793Z"},"links":{"cited_paper":"/paper/1803.03635","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:652b55d6f6552c6e2e4c1180a2b3502a38691a1c179507beac611b91d8a1c9ba","observation_id":"1413985e-a75b-42c4-b4dd-4eb5327cd5e8","resolution":{"observed_at":"2026-08-02T06:50:14.924793Z","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":"2026-08-02T06:50:14.927848Z","title":"A simple and effective pruning approach for large language models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.927848Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:f555cbdfe9fcb11d55bb6b708eb056dec178ebaa62607f3c7b35d28033d27d17","observation_id":"872934e3-2ea7-46e5-8f31-63e3bd86b54f","resolution":{"observed_at":"2026-08-02T06:50:14.927848Z","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":"2026-08-02T06:50:14.930824Z","title":"Llm-pruner: On the structural pruning of large language models.Advances in neural information processing systems, 36:21702–21720, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.930824Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:61cf7b9a6e0d913396bf2d3fd9c07a16e5063f5a058467af41a5869b9384e1ab","observation_id":"960cab5f-cbbc-416a-b0d1-6b2fef3c4e32","resolution":{"observed_at":"2026-08-02T06:50:14.930824Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.05162","last_updated":"2024-10-24T19:21:52Z","snapshot_observed_at":"2026-08-02T21:32:36.729604Z","submitted_at":"2024-02-07T18:34:38Z","title":"Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.05162","snapshot_observed_at":"2026-08-02T06:50:14.933979Z","title":"Assessing the brittleness of safety alignment via pruning and low-rank modifications.arXiv preprint arXiv:2402.05162, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.933979Z"},"links":{"cited_paper":"/paper/2402.05162","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:3ea0fe7036eb2cdb016b4bb396f16e790fa976af27fb062a68c2ad98f24c2040","observation_id":"bcc8117c-adf3-4adc-a3af-4b28f7f06d9f","resolution":{"observed_at":"2026-08-02T06:50:14.933979Z","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":"2026-08-02T06:50:14.936549Z","title":"Attention is all you need.Advances in neural information processing systems, 30, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.936549Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:4ed4e745b4417768b9a97054988a5de2dd09fa5011c15a89bf62ffbdf9c09adb","observation_id":"76edfb29-ef8c-4681-b385-613713670a63","resolution":{"observed_at":"2026-08-02T06:50:14.936549Z","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":"2026-08-02T06:50:14.938619Z","title":"Are sixteen heads really better than one?Advances in neural information processing systems, 32, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.938619Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:f3c4040db3827fd22a27b660de7ed718f6282921f1ca327bc66cfc4ec92ac7c1","observation_id":"8379b3b1-9257-4288-b01a-fc704ffb8998","resolution":{"observed_at":"2026-08-02T06:50:14.938619Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.21118","last_updated":"2024-11-04T02:08:55Z","snapshot_observed_at":"2026-07-06T18:54:41.705593Z","submitted_at":"2024-07-30T18:19:38Z","title":"Palu: Compressing KV-Cache with Low-Rank Projection","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.21118","snapshot_observed_at":"2026-08-02T06:50:14.941014Z","title":"Palu: Compressing kv-cache with low-rank projection.arXiv preprint arXiv:2407.21118, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.941014Z"},"links":{"cited_paper":"/paper/2407.21118","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:862a763e5c6d374ae63e0a9a5fc7b4a56b29c3ddd6f9d8d4bb097bb5eae1d980","observation_id":"7c90bfd3-a5db-4c70-8f38-0575b9e87fb7","resolution":{"observed_at":"2026-08-02T06:50:14.941014Z","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":"2026-08-02T06:50:14.944115Z","title":"When efficiency meets safety: A benchmark security analysis of kv cache compression in large language models","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.944115Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:046c4086beaa0400b37fefa49f22e02339a2e9af51614d1850d5260b0f47fe34","observation_id":"e246d9ea-7eb0-4ec5-8390-8d4c7fb5064b","resolution":{"observed_at":"2026-08-02T06:50:14.944115Z","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":"2026-08-02T06:50:14.946177Z","title":"Pruning for protection: Increasing jailbreak resistance in aligned llms without fine-tuning","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.946177Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:43a2cdab0e14f7255fbb28d1b516485eb67a29d8672deb66727ce8b6f971566b","observation_id":"9f5fa137-025a-4000-b0d2-b8dc23fe93fc","resolution":{"observed_at":"2026-08-02T06:50:14.946177Z","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":"2026-08-02T06:50:14.948403Z","title":"Edgeshard: Efficient llm inference via collaborative edge com- puting.IEEE Internet of Things Journal, 12(10):13119–13131, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.948403Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:13c0d79893101013d7b5f6426ffff7dfe7440c46c78ff3de8a1307574d42a9cd","observation_id":"e4b4fb82-c456-4066-b07d-2e513f04895a","resolution":{"observed_at":"2026-08-02T06:50:14.948403Z","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":"2026-08-02T06:50:14.950716Z","title":"Attacking and protecting data privacy in edge–cloud collaborative inference systems.IEEE Internet of Things Journal, 8(12):9706–9716, 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.950716Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:e20c85bf6e2f8ba0335378d8918bb78911d25a34a30219f278d335cb0211056f","observation_id":"d1ff58ae-9b9a-4a73-a191-d45db88330ba","resolution":{"observed_at":"2026-08-02T06:50:14.950716Z","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":"2026-08-02T06:50:14.953280Z","title":"Prompt inference attack on distributed large language model inference frameworks","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.953280Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:c35977093e154db5694097c1f2dd55f0e7089f55e62db862b76a65a12d5fa2a8","observation_id":"a8a1084d-7083-41ac-b12e-628a39063657","resolution":{"observed_at":"2026-08-02T06:50:14.953280Z","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":"2026-08-02T06:50:14.955579Z","title":"Algen: Few-shot inversion attacks on textual embeddings via cross-model alignment and generation","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.955579Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:c16f17c957a89025f5e8bc6cfe977d32294a2051831c52db0273d173ceb17238","observation_id":"7f373b04-6fea-4e77-87e3-e37235906883","resolution":{"observed_at":"2026-08-02T06:50:14.955579Z","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":"2026-08-02T06:50:14.957709Z","title":"Calibrating noise to sensitivity in private data analysis","venue":null,"work_id":null,"year":2006},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.957709Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:2edef500096bfb14b5bc2f976220df41825773488cb618f02a271ffca44eb613","observation_id":"1c2a75ed-3bc1-4a1b-a618-fef9c0aa9e26","resolution":{"observed_at":"2026-08-02T06:50:14.957709Z","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":"2026-08-02T06:50:14.959942Z","title":"Qlora: Efficient finetuning of quantized llms.Advances in neural information processing systems, 36:10088–10115, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.959942Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:8e5225e889283f504f160103d26664c87f8f6db82cc08155a80662ea1477c1f0","observation_id":"42e204af-43c3-492a-9c40-4ccb97241f7c","resolution":{"observed_at":"2026-08-02T06:50:14.959942Z","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":"2026-08-02T06:50:14.962477Z","title":"Lora: Low-rank adaptation of large language models.Iclr, 1(2):3, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.962477Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:7ffafc9388c423f50b0f73de34b38b2643749d8c4b35d916eb5ff2a1ea0ce0f9","observation_id":"4a251a6c-54e5-4e13-b3b3-1d9cd6d62972","resolution":{"observed_at":"2026-08-02T06:50:14.962477Z","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":"2026-08-02T06:50:14.965090Z","title":"Lora as oracle.arXiv preprint arXiv:2601.11207, 2026","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.965090Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:00f84b7480fb88d5fc1d165b27ffce2fe3755c50b036205e094f2b77cffafd50","observation_id":"b4fa1a3c-8115-45f0-97fb-ffcea8d72180","resolution":{"observed_at":"2026-08-02T06:50:14.965090Z","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":"2026-08-02T06:50:14.967390Z","title":"Membership inference attacks against machine learning models","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.967390Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:c5d30d7e3b5fdaed084676fae786459d97d1402e62a82e2923e7fa1a85153265","observation_id":"0092ce85-4f66-4a2e-bd4f-de31dfc399bf","resolution":{"observed_at":"2026-08-02T06:50:14.967390Z","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":"2026-08-02T06:50:14.969882Z","title":"Extracting training data from large language models.USENIX Security Symposium, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.969882Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:a5cedc726e0a6a00d2623dc00fc7197781fc775129b9c1919403a32047b03f4a","observation_id":"12cf418e-e97f-41a7-8a4a-2c5d08e676dc","resolution":{"observed_at":"2026-08-02T06:50:14.969882Z","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":"2026-08-02T06:50:14.972037Z","title":"Quantifying memorization across neural language models","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.972037Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:056287bdf168b53d3ba04d4bd14320fc83654a2ef2908dbfc86f557236b0a20d","observation_id":"afdf589a-7001-4f4e-a4fa-bcc8ea57b266","resolution":{"observed_at":"2026-08-02T06:50:14.972037Z","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":"2026-08-02T06:50:14.973954Z","title":"Morris Chang","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.973954Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:6defb246f244d280b50e094b201d88a87fa247e95c8780d6e8d4f5db7c1756c5","observation_id":"0bd3997a-6540-4d72-843e-f570b4e3bf0d","resolution":{"observed_at":"2026-08-02T06:50:14.973954Z","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":"2026-08-02T06:50:14.976040Z","title":"Federated fine-tuning of large language models under heterogeneous tasks and client resources.Advances in Neural Information Processing Systems, 37:14457–14483, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.976040Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:bbb13944b60d9ce4a9b1ff441098843b8d3a8c513fb9a942b7a6bc8621c8f203","observation_id":"ef8642be-d38a-4aff-8f21-4d3470192fa4","resolution":{"observed_at":"2026-08-02T06:50:14.976040Z","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":"2026-08-02T06:50:14.977958Z","title":"Deep leakage from gradients","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.977958Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:1aaad705a394cf9329b015b3abeb8177290a20a3a172244f672aa55e492d0f03","observation_id":"9d15a141-5233-4d50-8c22-d5457d594a32","resolution":{"observed_at":"2026-08-02T06:50:14.977958Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2508.02079","last_updated":"2025-08-04T05:45:24Z","snapshot_observed_at":"2026-07-06T22:07:16.776694Z","submitted_at":"2025-08-04T05:45:24Z","title":"AlignGuard-LoRA: Alignment-Preserving Fine-Tuning via Fisher-Guided Decomposition and Riemannian-Geodesic Collision Regularization","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2508.02079","snapshot_observed_at":"2026-08-02T06:50:14.980087Z","title":"Alignguard-lora: Alignment-preserving fine-tuning via fisher-guided de- composition and riemannian-geodesic collision regularization.arXiv preprint arXiv:2508.02079, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.980087Z"},"links":{"cited_paper":"/paper/2508.02079","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:2330fb48f6044b63cfe04ebc4bf798490e016c79fd4df4564ae62fcf513b25a3","observation_id":"ae42c21c-08ab-44bd-95f9-d9d81d119188","resolution":{"observed_at":"2026-08-02T06:50:14.980087Z","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":"2026-08-02T06:50:14.982394Z","title":"Retrieval-augmented generation for knowledge- intensive nlp tasks.Advances in neural information processing systems, 33:9459–9474, 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.982394Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:89aa2951fdbfe761325b51894bbc87a713cf46305b6641db9e3ddfd292ae8191","observation_id":"b83f6a36-3faa-49c0-be23-dbe3aa984f5c","resolution":{"observed_at":"2026-08-02T06:50:14.982394Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.18865","last_updated":"2024-02-29T05:27:45Z","snapshot_observed_at":"2026-07-06T17:37:13.985339Z","submitted_at":"2024-02-29T05:27:45Z","title":"Analyzing and Reducing Catastrophic Forgetting in Parameter Efficient Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.18865","snapshot_observed_at":"2026-08-02T06:50:14.984239Z","title":"Analyzing and reducing catastrophic forgetting in parameter efficient tuning.arXiv preprint arXiv:2402.18865, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.984239Z"},"links":{"cited_paper":"/paper/2402.18865","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:f66910b888d8bc9e5a38154b10dfe8a9dcdc0337b025acbaee451ce09ad66cf4","observation_id":"798acf14-67e8-491c-97d5-705bb7873ef1","resolution":{"observed_at":"2026-08-02T06:50:14.984239Z","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":"2026-08-02T06:50:14.986570Z","title":"Certified adversarial robustness via randomized smoothing","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.986570Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:04116d7232d3e9f4a755b2a1d1bb92e0d9ef51fda686f1032fb545e36ac57f21","observation_id":"1c386bdd-6e23-429f-9a72-7d91beed3539","resolution":{"observed_at":"2026-08-02T06:50:14.986570Z","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":"2026-08-02T06:50:14.989110Z","title":"What disease does this patient have? a large-scale open domain question answering dataset from medical exams.Applied Sciences, 11(14):6421, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.989110Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:a3c259430e98d886bcd103292ea62ad45deb9db912fc0a70c37518857c227c6f","observation_id":"14f5019f-ebf1-4edf-8e5b-c2ae4dff5514","resolution":{"observed_at":"2026-08-02T06:50:14.989110Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2009.03300","last_updated":"2021-01-12T18:57:11Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2020-09-07T17:59:25Z","title":"Measuring Massive Multitask Language Understanding","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2009.03300","snapshot_observed_at":"2026-08-02T06:50:14.991226Z","title":"Measuring massive multi- task language understanding.arXiv preprint arXiv:2009.03300, 2020","venue":null,"work_id":null,"year":2009},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.991226Z"},"links":{"cited_paper":"/paper/2009.03300","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:8a380bb92cfaf808f4f73a7d18f1067c04eadab5910658e07ff5fa085a0b6ac1","observation_id":"828c5745-a6bd-48a6-b295-bbc4312cd9ef","resolution":{"observed_at":"2026-08-02T06:50:14.991226Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2107.03374","last_updated":"2021-07-14T17:16:02Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2021-07-07T17:41:24Z","title":"Evaluating Large Language Models Trained on Code","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2107.03374","snapshot_observed_at":"2026-08-02T06:50:14.993630Z","title":"Evaluating large language models trained on code.arXiv preprint arXiv:2107.03374, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.993630Z"},"links":{"cited_paper":"/paper/2107.03374","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:a4b046ca7d539485bee0304c00e591b8a206809cda48517a57a0881024cb8480","observation_id":"d0dcf847-6034-4691-b985-2e791ab2b8ab","resolution":{"observed_at":"2026-08-02T06:50:14.993630Z","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":"2026-08-02T06:50:14.996408Z","title":"Edge-mpq: Layer-wise mixed-precision quantization with tightly integrated versatile inference units for edge computing","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.996408Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:d8177ef71397ab418b5300b8fa071b6bee52fdf548763cb0d46b7fc118aba491","observation_id":"100a7419-c880-4902-a5f0-46ee9b3a4b7f","resolution":{"observed_at":"2026-08-02T06:50:14.996408Z","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":"2026-08-02T06:50:14.998969Z","title":"Measuring massive multitask language understanding.Proceedings of the International Conference on Learning Representations (ICLR), 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:14.998969Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:d438288d89ee3cc26e293e85973866112cdf25dd71193f395b360d6e7385a57d","observation_id":"8962cff8-a341-4289-bb8b-2627367d7411","resolution":{"observed_at":"2026-08-02T06:50:14.998969Z","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":"2026-08-02T06:50:15.001295Z","title":"Measuring massive mul- titask language understanding (github repository)","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:15.001295Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:a3cf4d37c165ea3112a0d5848c0976987525acaa66700ece045e83f15db3dbec","observation_id":"26c56a31-2f03-48a2-a95f-19906133c62c","resolution":{"observed_at":"2026-08-02T06:50:15.001295Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.13833","last_updated":"2024-08-23T00:04:31Z","snapshot_observed_at":"2026-07-06T18:48:42.015848Z","submitted_at":"2024-07-18T18:06:59Z","title":"Phi-3 Safety Post-Training: Aligning Language Models with a \"Break-Fix\" Cycle","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.13833","snapshot_observed_at":"2026-08-02T06:50:15.003575Z","title":"Phi-3 safety post-training: Aligning language models with a” break-fix” cycle.arXiv preprint arXiv:2407.13833, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":58,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:15.003575Z"},"links":{"cited_paper":"/paper/2407.13833","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:c2cc99c9ba994e52e16a9d9a9d49e269c3ce5890e74f5f0d4d3488071aa5a45f","observation_id":"011e9db7-1450-42c9-b31c-a9428774f455","resolution":{"observed_at":"2026-08-02T06:50:15.003575Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2409.12186","last_updated":"2024-11-12T13:24:25Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-09-18T17:57:57Z","title":"Qwen2.5-Coder Technical Report","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2409.12186","snapshot_observed_at":"2026-08-02T06:50:15.006812Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":59,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:15.006812Z"},"links":{"cited_paper":"/paper/2409.12186","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:145395bcd7c8d9bef202b8629587206c395e0ff2d0a1bbb5ab2e41b1a5bdf255","observation_id":"f670b255-91ac-4750-bdf5-8fc1f7319787","resolution":{"observed_at":"2026-08-02T06:50:15.006812Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.00118","last_updated":"2024-10-02T15:22:49Z","snapshot_observed_at":"2026-08-02T16:20:09.773989Z","submitted_at":"2024-07-31T19:13:07Z","title":"Gemma 2: Improving Open Language Models at a Practical Size","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.00118","snapshot_observed_at":"2026-08-02T06:50:15.009992Z","title":"Gemma 2: Im- proving open language models at a practical size.arXiv preprint arXiv:2408.00118, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":60,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:15.009992Z"},"links":{"cited_paper":"/paper/2408.00118","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:b5b2bae6b49baa0fa33182bd63cfe23bd76ddd5ff5153c0dd764da55c2732a6f","observation_id":"702a0fdd-ec96-411a-99d2-82532e8027f9","resolution":{"observed_at":"2026-08-02T06:50:15.009992Z","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":"2026-08-02T06:50:15.012910Z","title":"Gemma 4 technical report, 2026","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":61,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:15.012910Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:7830f52079b64789b0532c29feb42ca580cc12c7bb110dd0c7757a9faa1808a3","observation_id":"031f8c2a-8541-4c99-b4fa-a7fb2983150e","resolution":{"observed_at":"2026-08-02T06:50:15.012910Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2407.21783","last_updated":"2024-11-23T23:27:33Z","snapshot_observed_at":"2026-07-06T18:55:11.576666Z","submitted_at":"2024-07-31T17:54:27Z","title":"The Llama 3 Herd of Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.21783","snapshot_observed_at":"2026-08-02T06:50:15.015602Z","title":"The llama 3 herd of models.arXiv preprint arXiv:2407.21783, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":62,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:15.015602Z"},"links":{"cited_paper":"/paper/2407.21783","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:dcc20ffccab26734c9a8ce505dd53bc240beefd90106a78c5b15d818eec56f7e","observation_id":"17129004-a3d0-41c4-bf9a-e9888be7792b","resolution":{"observed_at":"2026-08-02T06:50:15.015602Z","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":"2026-08-02T06:50:15.018313Z","title":"Language models are unsupervised multitask learners.OpenAI blog, 1(8):9, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":63,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:15.018313Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:d6998390c1a1ec25776e23746615da95b5b7342327a7da2e7a4bebfa71fe60fe","observation_id":"a41ea142-031b-4be5-b8c6-ad37ded71a54","resolution":{"observed_at":"2026-08-02T06:50:15.018313Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2403.15447","last_updated":"2024-06-04T05:40:12Z","snapshot_observed_at":"2026-07-06T17:49:11.762370Z","submitted_at":"2024-03-18T01:38:19Z","title":"Decoding Compressed Trust: Scrutinizing the Trustworthiness of Efficient LLMs Under Compression","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.15447","snapshot_observed_at":"2026-08-02T06:50:15.020939Z","title":"Decoding compressed trust: Scrutinizing the trustworthiness of efficient llms under compression.arXiv preprint arXiv:2403.15447, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":64,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:15.020939Z"},"links":{"cited_paper":"/paper/2403.15447","citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:9baf39e2694e849c33d84461a33ab4fa5fe58cbdbc30fd181094af40202ca9cd","observation_id":"0f4f0e1a-a9fc-4678-8c71-5b184f558fc6","resolution":{"observed_at":"2026-08-02T06:50:15.020939Z","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":"2026-08-02T06:50:13.359014Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge","version":1},"reference_index":2002,"source":"pdf_text","source_observed_at":"2026-08-02T06:50:13.359014Z"},"links":{"citing_paper":"/paper/2607.13088"},"observation_digest":"sha256:ea58867b13d32344144db9b036058325a4c4c30c1b83bd14b4ab805a07b840ef","observation_id":"4f8e46e0-ddb2-423c-ada0-afefc8b9dc17","resolution":{"observed_at":"2026-08-02T06:50:13.359014Z","resolver_source":null,"status":"parse_uncertain"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2607.13088","last_updated":"2026-07-13T16:45:04Z","latest_version":1,"primary_category":"cs.CR","snapshot_observed_at":"2026-08-02T21:33:36.438557Z","submitted_at":"2026-07-13T16:45:04Z","title":"Securing LLMs in the Wild: Privacy and Security Challenges at the Edge"},"reference_resolution":{"displayed":65,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":1,"unresolved":64,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":65},"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":"2026-08-03T06:30:56.289259+00:00","source":"crossref"},{"observed_at":"2026-08-03T06:30:50.922721+00:00","source":"retraction_watch"}],"thesis":"As of 3 August 2026, this Paper Citation Record lists 65 of 65 outbound references and 0 inbound Pith citation observations for arXiv:2607.13088."}