{"as_of":"2026-08-17T01:40:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:e727b75fac94133afe6bebdd0f15c6a188f435bfdc4a5cb8e811d65e8b4dceb2","coverage":[{"denominator":58,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":58,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-15T20:05:13.806297Z","state":"measured"},{"denominator":58,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":58,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-16T06:30:59.297886+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/2506.13494/citation-record","integrity":"/paper/2506.13494/integrity","json":"/paper/2506.13494/citation-record.json","paper":"/paper/2506.13494"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.498133Z","title":null,"venue":null,"work_id":"6e4205f4-b961-401c-876b-012ad438d9d7","year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.581105Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:fee3f431756b0da2421f34c13b4c1d34abb4ea27ec8a12f0a3ed0da6e6bbdffd","observation_id":"f21f770a-5e0a-4e3d-9752-2142a82234c7","resolution":{"observed_at":"2026-08-15T20:05:14.501622Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.488688Z","title":null,"venue":null,"work_id":"bfc38252-61e2-4255-a74f-e0d2345b9299","year":null},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.586157Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:59746072412770bb637440eac72c714389a07d9ed891be4aa19088f047d0d451","observation_id":"294c8eda-43c8-45bf-b16e-4d6a01436ba1","resolution":{"observed_at":"2026-08-15T20:05:14.491557Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.478562Z","title":null,"venue":null,"work_id":"0013440f-ae70-4d06-8848-c973bf299214","year":null},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.590572Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:053327fc1771db6a4689c471b4ea215f1d4f6c591f0af79a9e8cd7acf3536cbb","observation_id":"5e3f115b-32c4-45f9-a4bf-451414c47d55","resolution":{"observed_at":"2026-08-15T20:05:14.481842Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.467883Z","title":"Turning Your Weakness Into a Strength: Watermarking Deep Neural Networks by Backdooring","venue":null,"work_id":"fb860826-d155-41a3-b893-c3291fa83559","year":2018},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.594463Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:4429b85b4f1c599b0d49a91dde18218b813468bc06265ac67662abbb02ce3496","observation_id":"c19b11aa-276c-40b6-99ef-578d70ed0d40","resolution":{"observed_at":"2026-08-15T20:05:14.471644Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.457654Z","title":"Benchmarking Large Language Models in Retrieval- Augmented Generation","venue":null,"work_id":"b8c39fd0-df93-469a-a0a1-9e1b4c3cad91","year":null},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.598889Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:d612025dcd5ec49093fd8b7a35686a1105810b68d68c58869f982f49d29c5568","observation_id":"6df6c065-0643-42fb-bdde-86fa61f206a5","resolution":{"observed_at":"2026-08-15T20:05:14.461336Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.447589Z","title":"BadNL: Backdoor Attacks Against NLP Models with Semantic-preserving Improvements","venue":null,"work_id":"f1fb8ca6-d62e-406d-ade6-d128222a1beb","year":2021},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.602594Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:949becd79ec2d9ff03b7b5de9624f46504383d922b4de456af9d16380288404d","observation_id":"fc60c94e-7384-4928-b6cd-49388e42d49a","resolution":{"observed_at":"2026-08-15T20:05:14.451137Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.436513Z","title":"REFIT: A Unified Watermark Removal Framework For Deep Learning Systems With Limited Data","venue":null,"work_id":"2750c5ad-3b50-4d0f-a3a5-72783a28116e","year":2021},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.606789Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:464c709e874efd24171019a1b6746e35f80e5058ac985aaab097167d131e4938","observation_id":"774ab2c9-0374-48a4-9c2c-2baff3bc92d2","resolution":{"observed_at":"2026-08-15T20:05:14.440403Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2105.06762","last_updated":"2021-06-16T06:34:42Z","snapshot_observed_at":"2026-08-16T18:24:45.131069Z","submitted_at":"2021-05-14T11:12:40Z","title":"DialogSum: A Real-Life Scenario Dialogue Summarization Dataset","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2105.06762","snapshot_observed_at":"2026-08-15T20:05:13.610432Z","title":"DialogSum: A Real-Life Scenario Dialogue Summa- rization Dataset","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.610432Z"},"links":{"cited_paper":"/paper/2105.06762","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:9ca47782e868f7f443f23a28b13a97e736c0a901908830e547510af1bb3a8820","observation_id":"6f0ff7fa-4da7-4d74-b59b-41dfafd005db","resolution":{"observed_at":"2026-08-15T20:05:13.610432Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.425571Z","title":"Increasing Diversity While Maintaining Ac- curacy: Text Data Generation with Large Language Models and Human Interventions","venue":null,"work_id":"fcdb4c0b-a06d-40d5-a5ed-c3a4bd9864eb","year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.614411Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:e663cd3ba09bee83ed6359abd4ecc7e16d1015f999d7c0de3235d605388d09ae","observation_id":"88fd7c3b-c086-431f-b272-11eaaff58099","resolution":{"observed_at":"2026-08-15T20:05:14.429454Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.416012Z","title":"SSL- Guard: A Watermarking Scheme for Self-supervised Learning Pre-trained Encoders","venue":null,"work_id":"a67cc1f0-194b-45fe-a9bc-a2f35a457f99","year":2022},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.618625Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:238bec6d4dfce7091baed6a118ac29034c5253b5197401eb2df88280d085b5a3","observation_id":"af7e8229-016b-4afa-ac9d-c8e2815257c4","resolution":{"observed_at":"2026-08-15T20:05:14.419129Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.407061Z","title":"BERT: Pre-training of Deep Bidi- rectional Transformers for Language Understanding","venue":null,"work_id":"69a73e4c-e7cf-446d-8f70-933764ab35a3","year":2019},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.622335Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:e6a929f452179f993d27d789af008f2ea6575d18c0b2ee1a4607ec9f438aeed5","observation_id":"56e007be-c73a-464c-8c80-590b81d401a1","resolution":{"observed_at":"2026-08-15T20:05:14.410000Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.396001Z","title":"Watermark Removal Scheme Based on Neural Network Model Pruning","venue":null,"work_id":"3aa36180-d92a-46db-bac8-9f37ea5174f1","year":2022},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.625975Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:7ba289470163fc9f890b088e9b7c8a3e31b038cc5ed350670b97d877a4eaf889","observation_id":"ec1dbe64-129e-437f-ab22-bbb99306cbfb","resolution":{"observed_at":"2026-08-15T20:05:14.399935Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.384757Z","title":"Fine-tuning Is Not Enough: A Simple yet Effective Watermark Removal Attack for DNN Models","venue":null,"work_id":"8e4627be-36b3-4e83-b2e8-a708e2c1369c","year":null},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.629634Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:6ef1f339df53cc81275362612ec6a35a7aeaad2ab8aacc9b616b275c6b8221f9","observation_id":"a8478f51-98b2-4f98-8bda-39ce046cf39b","resolution":{"observed_at":"2026-08-15T20:05:14.388865Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.373406Z","title":"Choquette-Choo, Varun Chandrasekaran, and Nicolas Papernot","venue":null,"work_id":"43b797aa-9800-45aa-b204-cbe9a63e2606","year":1937},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.634292Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:cb9e1e70c47207e7bfec5f455d5d2dfb3ddbce7c4fc82cab8f300e7190616ee8","observation_id":"ece92ced-ce9c-40cd-93c3-ee64ab2a5edd","resolution":{"observed_at":"2026-08-15T20:05:14.377407Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.19361","last_updated":"2024-06-24T14:48:29Z","snapshot_observed_at":"2026-08-16T14:14:03.069592Z","submitted_at":"2024-02-29T17:12:39Z","title":"Watermark Stealing in Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.19361","snapshot_observed_at":"2026-08-15T20:05:13.641762Z","title":"Watermark Stealing in Large Language Models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.641762Z"},"links":{"cited_paper":"/paper/2402.19361","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:8f05a6c520b2a11a18b0428cf97507f094b3d78d6195bdfd82857deb578a06c5","observation_id":"877653c0-1ad8-4410-bbe0-deecc15ca846","resolution":{"observed_at":"2026-08-15T20:05:13.641762Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.362034Z","title":"A Watermark for Large Language Models","venue":null,"work_id":"5f6ac72a-7378-4eba-bbb0-74d67c1db1b0","year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.646119Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:3958519646fa226754407b328daff7346a687b8297bff901eaffb9ba7913ddea","observation_id":"3b370c68-d58e-44f9-ba5f-25ae2a105d08","resolution":{"observed_at":"2026-08-15T20:05:14.365864Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2306.04634","last_updated":"2024-05-01T21:20:36Z","snapshot_observed_at":"2026-08-16T15:25:44.546287Z","submitted_at":"2023-06-07T17:58:48Z","title":"On the Reliability of Watermarks for Large Language Models","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2306.04634","snapshot_observed_at":"2026-08-15T20:05:13.649880Z","title":"On the Reliability of Watermarks for Large Language Models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.649880Z"},"links":{"cited_paper":"/paper/2306.04634","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:caa1faf71bbfcd75d990233cba2844260a794880a3a37c06f22eaeb674fdedd9","observation_id":"c7d9d5f9-7355-4407-ba33-36326f4ee8e9","resolution":{"observed_at":"2026-08-15T20:05:13.649880Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.351106Z","title":"Large Lan- guage Models are Zero-Shot Reasoners","venue":null,"work_id":"ca8298f6-891e-4cf8-9638-924153988581","year":2022},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.653752Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:703e9d57458331a64221a46213fb459b4a61e969723cfffff23806266bc61483","observation_id":"6f3cb3ad-6ad4-4fd9-966d-496e4b041af5","resolution":{"observed_at":"2026-08-15T20:05:14.354883Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2309.06180","last_updated":"2023-09-12T12:50:04Z","snapshot_observed_at":"2026-08-02T09:51:08.145755Z","submitted_at":"2023-09-12T12:50:04Z","title":"Efficient Memory Management for Large Language Model Serving with PagedAttention","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2309.06180","snapshot_observed_at":"2026-08-15T20:05:13.658014Z","title":"Efficient Memory Management for Large Language Model Serving with PagedAttention","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.658014Z"},"links":{"cited_paper":"/paper/2309.06180","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:fe58edcd79ed77c47b66ce2bba53a64ecdba5ff173883eb41b63c4881089626b","observation_id":"1dd7d24e-7d69-48e8-98c3-b6cf94fc0119","resolution":{"observed_at":"2026-08-15T20:05:13.658014Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.340542Z","title":"Who Wrote this Code? Watermarking for Code Generation","venue":null,"work_id":"deac08e2-4db4-4aeb-b44c-9280eb2733df","year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.662335Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:62efb477d0bc15c0cd1015dcc32e1e46d5cce87d2ab24d84139278bf31192f32","observation_id":"e2810871-7c1a-4425-bab5-c17a23e03708","resolution":{"observed_at":"2026-08-15T20:05:14.343826Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.331512Z","title":"PLMmark: A Se- cure and Robust Black-Box Watermarking Framework for Pre-trained Language Models","venue":null,"work_id":"31d0d84d-95cd-4851-8a25-bcaa6197c915","year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.666229Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:507af58174b8bff7a47612f70acbfa9774eaf42aee0214a45f28c1b2e3ea1313","observation_id":"836de15c-2712-4e82-becc-cb8fdf2d9c94","resolution":{"observed_at":"2026-08-15T20:05:14.334513Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.321584Z","title":"Synthetic Data Generation with Large Language Models for Text Classification: Potential and Limita- tions","venue":null,"work_id":"9ea4b576-7dfa-4697-9a20-5436f17cd35b","year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.669940Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:c627ec649bf235629961500736953bb802d8ee1126e3e37bd8f38c35934b46d3","observation_id":"db8176b6-8302-4cad-89b0-934a440747f3","resolution":{"observed_at":"2026-08-15T20:05:14.325164Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2404.01268","last_updated":"2024-04-01T17:45:15Z","snapshot_observed_at":"2026-08-16T22:36:56.436417Z","submitted_at":"2024-04-01T17:45:15Z","title":"Mapping the Increasing Use of LLMs in Scientific Papers","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2404.01268","snapshot_observed_at":"2026-08-15T20:05:13.673700Z","title":"Manning, and James Y","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.673700Z"},"links":{"cited_paper":"/paper/2404.01268","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:02a839cbc21a3ca7d80a9a77561f8befb4afb98cd1ecc1b9a1889f48e53e7ba2","observation_id":"1cf380d7-8840-47db-87cd-04d15168c19a","resolution":{"observed_at":"2026-08-15T20:05:13.673700Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.309340Z","title":"A Semantic Invariant Robust Watermark for Large Language Models","venue":null,"work_id":"3e71ba26-0494-4809-b09b-09b3eacc304a","year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.678029Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:d85eb2fbfbfc808fed7d7423aa808630f8010a3deb8e67ebb83673d258fe8c02","observation_id":"5709cfbd-2ad9-4e83-bea5-9bbf95ab70ad","resolution":{"observed_at":"2026-08-15T20:05:14.314053Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2310.03744","last_updated":"2024-05-15T19:22:44Z","snapshot_observed_at":"2026-08-15T02:35:59.111911Z","submitted_at":"2023-10-05T17:59:56Z","title":"Improved Baselines with Visual Instruction Tuning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.03744","snapshot_observed_at":"2026-08-15T20:05:13.681598Z","title":"Improved Baselines with Visual Instruction Tun- ing","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.681598Z"},"links":{"cited_paper":"/paper/2310.03744","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:47a345eaa15adb9f733d4fdd8f82465bf4cb6b34e34cc273ea4d75c89dc2619f","observation_id":"d040c723-74a4-4a6a-898d-59b7abaeabcb","resolution":{"observed_at":"2026-08-15T20:05:13.681598Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.298520Z","title":"Fine-Pruning: Defending Against Backdooring At- tacks on Deep Neural Networks","venue":null,"work_id":"bed265cc-0de3-4ca3-9b14-6922ce32acb1","year":null},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.686129Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:ca95b160e010bd7e7500dd4ecf61af0bfd151fb190038277f463f3e52f1c62d9","observation_id":"2b6d8041-2ee6-4f56-8685-68de15c1ea5d","resolution":{"observed_at":"2026-08-15T20:05:14.302410Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1907.11692","last_updated":"2019-07-26T17:48:29Z","snapshot_observed_at":"2026-08-16T14:33:50.657682Z","submitted_at":"2019-07-26T17:48:29Z","title":"RoBERTa: A Robustly Optimized BERT Pretraining Approach","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1907.11692","snapshot_observed_at":"2026-08-15T20:05:13.693923Z","title":"RoBERTa: A Ro- bustly Optimized BERT Pretraining Approach","venue":null,"work_id":null,"year":1907},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.693923Z"},"links":{"cited_paper":"/paper/1907.11692","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:c1616d06c443bf986ee3b8879732af19adccad001fe8cf437394728669a5df28","observation_id":"a72c2e65-4f76-4a6f-98fb-7cae0e049107","resolution":{"observed_at":"2026-08-15T20:05:13.693923Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:13.698022Z","title":"Robustness Over Time: Understanding Adversarial Examples’ Effective- ness on Longitudinal Versions of Large Language Mod- els","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.698022Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:478c0c8055db467ceeacdbc157108ada9727e97b07478f3ecb50481d8ef7649a","observation_id":"3474027f-9fcc-4880-8ed7-fb9e9ab6415f","resolution":{"observed_at":"2026-08-15T20:05:13.698022Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2301.01197","last_updated":"2023-01-03T16:58:34Z","snapshot_observed_at":"2026-08-16T16:04:38.848674Z","submitted_at":"2023-01-03T16:58:34Z","title":"Backdoor Attacks Against Dataset Distillation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2301.01197","snapshot_observed_at":"2026-08-15T20:05:13.701832Z","title":"Backdoor Attacks Against Dataset Distil- lation","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.701832Z"},"links":{"cited_paper":"/paper/2301.01197","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:91bb0085f433e0f488f6a5bb88e917021ca9b2160019ecb325d9d898ccc25a45","observation_id":"10c4be50-fcb6-4475-9dbd-659d3484340a","resolution":{"observed_at":"2026-08-15T20:05:13.701832Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.12502","last_updated":"2023-05-21T16:37:50Z","snapshot_observed_at":"2026-08-16T15:31:20.244368Z","submitted_at":"2023-05-21T16:37:50Z","title":"Watermarking Diffusion Model","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.12502","snapshot_observed_at":"2026-08-15T20:05:13.705252Z","title":"Watermarking Diffusion Model","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.705252Z"},"links":{"cited_paper":"/paper/2305.12502","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:e322b0dadb88dbfe7cb1015b72ef040977cb94e125b9bfebec25d747c8dc4b5f","observation_id":"f50ef0e8-70fc-4dd9-b5f6-e332d0090b9b","resolution":{"observed_at":"2026-08-15T20:05:13.705252Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.276600Z","title":"SoK: How Robust is Image Classification Deep Neural Network Watermarking? In IEEE Sympo- sium on Security and Privacy (S&P)","venue":null,"work_id":"60d96bf7-ad28-4d53-880a-72a63bd44450","year":2022},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.708808Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:4c77a8e84abbed79a688005db0e0c92fd269043e2b3962f23fb2aa7dd8fcce2b","observation_id":"c354a406-f805-408c-a260-2b2546571b17","resolution":{"observed_at":"2026-08-15T20:05:14.280560Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.266063Z","title":"Maas, Raymond E","venue":null,"work_id":"48c8a827-679f-4ec9-9cea-14db94b66fa3","year":2011},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.711999Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:bbc224fd145047f64a0e807fdf77d129fb4177a57b19e60122d6c7f784b6353b","observation_id":"dcc0d5b4-0601-453e-bbcc-9cb077d8fe75","resolution":{"observed_at":"2026-08-15T20:05:14.269857Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1711.01894","last_updated":"2019-08-07T07:22:53Z","snapshot_observed_at":"2026-08-16T23:06:02.819404Z","submitted_at":"2017-11-06T13:57:08Z","title":"Adversarial Frontier Stitching for Remote Neural Network Watermarking","version":2},"cited_work":{"arxiv_id":"1711.01894","doi":null,"metadata_source":"pith","pith_arxiv_id":"1711.01894","snapshot_observed_at":"2026-08-15T20:05:13.919346Z","title":"Adversarial Frontier Stitching for Remote Neural Network Watermarking","venue":"cs.CR","work_id":"a690bd35-b27b-43ff-bea1-a310b0e7d3a9","year":2017},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.715549Z"},"links":{"cited_paper":"/paper/1711.01894","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:76a6bf74b291acde7a51cc2daa0e83adda47d156bda4ee0557759426af329d2a","observation_id":"0e7d8361-3dd8-4824-921c-9facf154e29f","resolution":{"observed_at":"2026-08-15T20:05:13.926416Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.255026Z","title":"Protecting Intellectual Property of Generative Adversarial Networks From Ambiguity Attacks","venue":null,"work_id":"bf25494c-61e2-468a-a5ae-b80346b2997e","year":null},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.719616Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:47814626745b29b5b3f8a8e63677d46e39d09e00f70d76937e83d06cead9a1f1","observation_id":"bfcd9b10-9cf4-498e-a29a-7b7d7eb7bf79","resolution":{"observed_at":"2026-08-15T20:05:14.259070Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2303.08774","last_updated":"2024-03-04T06:01:33Z","snapshot_observed_at":"2026-08-16T19:40:28.523700Z","submitted_at":"2023-03-15T17:15:04Z","title":"GPT-4 Technical Report","version":6},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.08774","snapshot_observed_at":"2026-08-15T20:05:13.723763Z","title":"GPT-4 Technical Report","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.723763Z"},"links":{"cited_paper":"/paper/2303.08774","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:aa69d38580ee8af7b5fcc7014dedd84f4275f2ec2f8da43154ea307d41ae0bea","observation_id":"90fc52f7-a726-444e-a1e3-a9fcbbe55770","resolution":{"observed_at":"2026-08-15T20:05:13.723763Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09731","last_updated":"2023-05-16T18:05:19Z","snapshot_observed_at":"2026-08-16T15:32:34.094615Z","submitted_at":"2023-05-16T18:05:19Z","title":"What In-Context Learning \"Learns\" In-Context: Disentangling Task Recognition and Task Learning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09731","snapshot_observed_at":"2026-08-15T20:05:13.727499Z","title":"What In-Context Learning \"Learns\" In-Context: Dis- entangling Task Recognition and Task Learning.CoRR abs/2305.09731, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.727499Z"},"links":{"cited_paper":"/paper/2305.09731","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:909661290e95cf0cbd2847ae12689b4e9adbfbf51d6df09c10ec226fd54e9826","observation_id":"614a7207-2f59-4f95-ae2a-6d002ea92906","resolution":{"observed_at":"2026-08-15T20:05:13.727499Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.242798Z","title":"Hidden Trigger Backdoor Attack on NLP Models via Linguistic Style Manipulation","venue":null,"work_id":"932b0d38-aaa3-4562-898e-f21d99fcd5e2","year":2022},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.731645Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:b4e8c6e813822eb4ba214f09d19ac7b8f7bf6e6dd8fc17f373db89a5419ceee6","observation_id":"f9ad02e8-5cd9-4a23-832e-55db64dde343","resolution":{"observed_at":"2026-08-15T20:05:14.246216Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.16187","last_updated":"2024-11-13T15:14:38Z","snapshot_observed_at":"2026-08-16T14:15:30.337020Z","submitted_at":"2024-02-25T20:24:07Z","title":"No Free Lunch in LLM Watermarking: Trade-offs in Watermarking Design Choices","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.16187","snapshot_observed_at":"2026-08-15T20:05:13.735365Z","title":"Attacking LLM Watermarks by Exploiting Their Strengths","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.735365Z"},"links":{"cited_paper":"/paper/2402.16187","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:5ba9fd96d4d8d4342ff08ae811c73e2e4229f607585c255b4a2cf1294e8e6c75","observation_id":"be0a15fe-e27e-4156-80c2-caba2d995a77","resolution":{"observed_at":"2026-08-15T20:05:13.735365Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.231747Z","title":"Can Large Language Models Rea- son about Program Invariants? In International Con- ference on Machine Learning (ICML)","venue":null,"work_id":"97c194bf-7d57-407f-9f26-2c0a20303d88","year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.739346Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:baf7ff004a0fd37b25c0382d18e7608029fc2220b4a82446b1a6a0b420bce5e1","observation_id":"19d1feb2-7c0b-42cb-a7ae-2ae20534a9a6","resolution":{"observed_at":"2026-08-15T20:05:14.235516Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.221042Z","title":"Are You Copying My 16 Model? Protecting the Copyright of Large Language Models for EaaS via Backdoor Watermark","venue":null,"work_id":"56b67953-e9c0-41b3-9f0f-49c66b2f5390","year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.742709Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:1f395b3300b7161dc7a22613e8e7c318b2ea714b1e9975e866e2b3dbd5339ccd","observation_id":"f5cfd16e-5c8f-40f8-88dc-f78535ddc13a","resolution":{"observed_at":"2026-08-15T20:05:14.224841Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.210359Z","title":"MAUVE: Measuring the Gap Between Neural Text and Human Text using Divergence Fron- tiers","venue":null,"work_id":"df2d4f26-75f7-41b2-975c-1968a8ee1f50","year":2021},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.746443Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:d8fa5fce61c157fb878b75e165e1497829c790274a63ff89626d5d3343dc46aa","observation_id":"a5f497f4-ec59-4027-8f15-a0d734377988","resolution":{"observed_at":"2026-08-15T20:05:14.214225Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.198826Z","title":null,"venue":null,"work_id":"6512ea1b-a25f-44a0-8389-10dc7ea3ec1e","year":2020},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.751009Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:bc84f71b25ef55219602730f81d671afe494e7b9cbb74886dad0f030e671b1ae","observation_id":"b377f731-04e0-4c54-b286-2852a122e48c","resolution":{"observed_at":"2026-08-15T20:05:14.202873Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2311.08721","last_updated":"2024-04-01T17:44:19Z","snapshot_observed_at":"2026-08-16T14:43:12.028399Z","submitted_at":"2023-11-15T06:19:02Z","title":"A Robust Semantics-based Watermark for Large Language Model against Paraphrasing","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2311.08721","snapshot_observed_at":"2026-08-15T20:05:13.754604Z","title":"A Ro- bust Semantics-based Watermark for Large Language Model against Paraphrasing","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.754604Z"},"links":{"cited_paper":"/paper/2311.08721","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:2bfad3cfb91880b263d808e22aea82e2a202efef8d124e1202d384f5844c7e98","observation_id":"d8b16b26-6ce3-487b-955f-d0137e1e9388","resolution":{"observed_at":"2026-08-15T20:05:13.754604Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1804.00750","last_updated":"2018-05-31T23:57:59Z","snapshot_observed_at":"2026-08-14T19:29:57.020152Z","submitted_at":"2018-04-02T22:23:04Z","title":"DeepSigns: A Generic Watermarking Framework for IP Protection of Deep Learning Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1804.00750","snapshot_observed_at":"2026-08-15T20:05:13.758607Z","title":"DeepSigns: A Generic Watermarking Framework for IP Protection of Deep Learning Mod- els","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.758607Z"},"links":{"cited_paper":"/paper/1804.00750","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:211ef2dd7766b92240d75e3287bf4f4b3e7fa6a8ecb51092b70fe2f664ffae54","observation_id":"ecf92461-5a14-4a0d-84cf-e9ddfef10b5b","resolution":{"observed_at":"2026-08-15T20:05:13.758607Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.187561Z","title":"Embedding Watermarks into Deep Neural Networks","venue":null,"work_id":"3b329c23-7c12-411a-a5d6-7f90fd45b859","year":2017},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.762917Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:7a33e86d4d8b15b3ddefbb53e15bc4f343675f75690681a91d04e339edbc4506","observation_id":"4e038ada-5af4-43ac-a73e-ecc55d8aa876","resolution":{"observed_at":"2026-08-15T20:05:14.190816Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.177047Z","title":"Attacks on Digital Watermarks for Deep Neural Networks","venue":null,"work_id":"4101469b-a347-4b5f-ad8a-51302067ccca","year":2019},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.766580Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:70f8496ea3e3b1beb71b7736dde63551bd72e818860923eb81ad4b39bc0a7159","observation_id":"26ae3b52-3151-45fd-922c-db8fb6306290","resolution":{"observed_at":"2026-08-15T20:05:14.180912Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.164941Z","title":"Chi, Quoc V","venue":null,"work_id":"88f48bc4-4687-4074-ae0b-e11fa1206af1","year":2022},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.769524Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:839daa18e7f95a3c4649ff39229f1cce530c98e680ecda89a9417a155c15b6dd","observation_id":"3cb94ae6-7c65-4d89-9f35-da10ce05c7b4","resolution":{"observed_at":"2026-08-15T20:05:14.169288Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2407.10671","last_updated":"2024-09-10T13:25:53Z","snapshot_observed_at":"2026-08-16T18:57:50.930866Z","submitted_at":"2024-07-15T12:35:42Z","title":"Qwen2 Technical Report","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2407.10671","snapshot_observed_at":"2026-08-15T20:05:13.772794Z","title":"Qwen2 Technical Report","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.772794Z"},"links":{"cited_paper":"/paper/2407.10671","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:714da2e46ebe98633b7330f9826cc20cc40a771cbd97d7d308579106984ae25f","observation_id":"b765b2b0-a7d7-48d1-b9bc-540dcbd4e803","resolution":{"observed_at":"2026-08-15T20:05:13.772794Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.15115","last_updated":"2025-01-03T02:18:21Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-12-19T17:56:09Z","title":"Qwen2.5 Technical Report","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.15115","snapshot_observed_at":"2026-08-15T20:05:13.776333Z","title":"Qwen2.5 Technical Report","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.776333Z"},"links":{"cited_paper":"/paper/2412.15115","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:8d353a8fd6b2ffe955563ed7e89a66a714ab6cef516d10098a3ac113eb1308ed","observation_id":"c964d4b0-05eb-47a1-867b-bd2f5b2e28ec","resolution":{"observed_at":"2026-08-15T20:05:13.776333Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.153464Z","title":"Stoecklin, Heqing Huang, and Ian Molloy","venue":null,"work_id":"2077a368-d2ab-4151-91d8-5f320f824fde","year":2018},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.779911Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:89543bf5300162d5b95de45c5d87d44df5ea959af8cf1a2d64bcfe3cd56697d1","observation_id":"42be5da1-8f5f-4cf2-a655-9585ca2a4fda","resolution":{"observed_at":"2026-08-15T20:05:14.157727Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.141843Z","title":"Instruction Backdoor Attacks Against Cus- tomized LLMs","venue":null,"work_id":"c16ffeec-1e34-4eb6-97bc-70babcec9cc1","year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.783366Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:0d0bd51a211044cb638ae9610a5a4e3e0f5b2b4d1935f2b270f82b057bc5a47e","observation_id":"473fa5af-2c68-4bc6-b097-28e9e56a59f5","resolution":{"observed_at":"2026-08-15T20:05:14.145588Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.128924Z","title":"Character-level Convolutional Networks for Text Clas- sification","venue":null,"work_id":"a6b285ad-3a6e-41c8-9cc9-69965c9d9620","year":null},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.787090Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:4dec98b08c7e54392f74902be672cc5408da16721479e535d270cdb362b07180","observation_id":"b95b7335-8f4d-4547-8b2f-a9bcadf61973","resolution":{"observed_at":"2026-08-15T20:05:14.132555Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.118432Z","title":"Provable Robust Watermarking for AI-Generated Text","venue":null,"work_id":"4f0fc36d-aca8-4c0f-a45c-60c54af52a77","year":2024},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.790806Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:43d66fdc55683a0056b0d94f6c6a9141a8864992ea8bf0ab7c91c220f46a09b4","observation_id":"b1acfadc-23f9-4dff-812b-c4c29bd25581","resolution":{"observed_at":"2026-08-15T20:05:14.122309Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.106672Z","title":"Attention Distrac- tion: Watermark Removal Through Continual Learn- ing with Selective Forgetting","venue":null,"work_id":"1dfab01f-4cc3-4d18-b4ee-c92a09a5426c","year":2022},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.794540Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:9b65ca100b892f5dab2181c993c0c7818f3e14fd88e8329ee0354abd56ac61dc","observation_id":"17b42f32-6dc5-4e3f-9f4b-d62e47324d90","resolution":{"observed_at":"2026-08-15T20:05:14.110614Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.095368Z","title":"Large Language Models are Human-Level Prompt En- gineers","venue":null,"work_id":"a5b9e02b-fc98-468b-bc47-f018dd9b8bd5","year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.798814Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:9711dce3852181704e759daab73dc259c5016c2d68de295008a8f461355947d1","observation_id":"309e7638-2de3-440d-bfb6-9128bdd7da02","resolution":{"observed_at":"2026-08-15T20:05:14.099059Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2304.10592","last_updated":"2023-10-02T16:38:35Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-04-20T18:25:35Z","title":"MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2304.10592","snapshot_observed_at":"2026-08-15T20:05:13.802572Z","title":"MiniGPT-4: Enhancing Vision- Language Understanding with Advanced Large Lan- guage Models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.802572Z"},"links":{"cited_paper":"/paper/2304.10592","citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:da694f0163d0a7aefbc3e512697d92a9be4c7d2c3d86447e1ab8dc3b9504dc6a","observation_id":"2e3487eb-5544-4941-8cc8-a28470d33ec7","resolution":{"observed_at":"2026-08-15T20:05:13.802572Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.084738Z","title":"To Prune, or Not to Prune: Exploring the Efficacy of Pruning for Model Compression","venue":null,"work_id":"9069ca87-fd0c-41a3-82e1-04e2bc4a0e0d","year":2018},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.806297Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:1f8efccee5aec34e5dd1a6de9beb7d993efca00cfc39292579bddfdf8e30cd71","observation_id":"d46fb075-40ed-4356-8f5c-66dd6a7fda23","resolution":{"observed_at":"2026-08-15T20:05:14.088526Z","resolver_source":"raw_fallback","status":"malformed_identifier"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-15T20:05:14.287808Z","title":null,"venue":null,"work_id":"b74bbfc3-e029-4669-a0cc-200f349b4d50","year":2018},"citing_paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks","version":1},"reference_index":294,"source":"pdf_text","source_observed_at":"2026-08-15T20:05:13.690312Z"},"links":{"citing_paper":"/paper/2506.13494"},"observation_digest":"sha256:5555fcd888f980d546aebc3390c1b5f9cbf76da2a0fdfa8b8f5684b5220c2315","observation_id":"13b20512-a735-4a35-9216-660dd4aaf8b1","resolution":{"observed_at":"2026-08-15T20:05:14.291279Z","resolver_source":"raw_fallback","status":"parse_uncertain"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-16T06:30:59.297886+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2506.13494","last_updated":"2025-06-16T13:51:49Z","latest_version":1,"primary_category":"cs.CR","snapshot_observed_at":"2026-08-15T19:57:52.769111Z","submitted_at":"2025-06-16T13:51:49Z","title":"Watermarking LLM-Generated Datasets in Downstream Tasks"},"reference_resolution":{"displayed":58,"state_counts":{"malformed_identifier":1,"metadata_mismatch":0,"parse_uncertain":1,"unresolved":22,"verified_exact":1,"verified_fuzzy":33},"total_outbound_references":58},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-16T06:30:59.297886+00:00","source":"crossref"},{"observed_at":"2026-08-16T06:30:54.164669+00:00","source":"retraction_watch"}],"thesis":"As of 17 August 2026, this Paper Citation Record lists 58 of 58 outbound references and 0 inbound Pith citation observations for arXiv:2506.13494."}