{"as_of":"2026-08-13T10:58:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:72a46fd29739c17c986d007a33fd50672f8871de19ae29912110cdda66204f66","coverage":[{"denominator":20,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":20,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-12T14:07:18.363722Z","state":"measured"},{"denominator":20,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":20,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-13T06:32:02.005865+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/2411.15673/citation-record","integrity":"/paper/2411.15673/integrity","json":"/paper/2411.15673/citation-record.json","paper":"/paper/2411.15673"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-12T14:07:18.707362Z","title":"V ATT: trans- formers for multimodal self-supervised learning from raw video, audio and text","venue":null,"work_id":"d8bd0675-e642-4c7d-b576-85c0355eb8f4","year":2021},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.270347Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:8f5467194e671a532ce6076ceed098ab6effb940d855f1da7fff1a6c8f22041d","observation_id":"22cc8354-9241-43cb-b650-826a485d67b0","resolution":{"observed_at":"2026-08-12T14:07:18.715323Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.623394Z","title":"Visual language maps for robot navigation","venue":null,"work_id":"ca92c7b9-d2af-40c3-9359-d495f919f78e","year":2023},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.292857Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:2c83e68ed83a6ae74598c41cfe003abd14ea42720d19c8d1af344324e44b5d56","observation_id":"320f684f-ed65-414d-8b29-67f78e1d4313","resolution":{"observed_at":"2026-08-12T14:07:18.627757Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.606078Z","title":"Align be- fore fuse: Vision and language representation learning with momentum distillation","venue":null,"work_id":"f6f2d45c-c22f-4f94-b24d-ded6fb8e0fd7","year":2022},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.299433Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:7f2450800e805b170f73f8186e9d4750bbe216c5ad8c49ab19b4b0b6ab36bf3a","observation_id":"d4b8fff7-75ca-448b-8911-79d26e56e83c","resolution":{"observed_at":"2026-08-12T14:07:18.611384Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.580665Z","title":"Backdoor defense with ma- chine unlearning","venue":null,"work_id":"60215986-f3a1-4354-84b0-02de1066d053","year":2022},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.309617Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:56a17e4a30b8de647b328510d830e5e5bfa218dd2082ae506a92cdbf2558c38e","observation_id":"c2f1ac1c-45e3-4ff5-8eaf-a687ed8100d6","resolution":{"observed_at":"2026-08-12T14:07:18.585104Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.562587Z","title":"Visual classification via de- scription from large language models","venue":null,"work_id":"4e4ede4d-2191-4294-8afc-993efb2e1274","year":2023},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.314385Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:2efc70cbe4251b0f62d42ab43e3237875a22b0233dcdb47cae44776276166b08","observation_id":"1bf2b972-e3fa-492d-830f-284bda07ea97","resolution":{"observed_at":"2026-08-12T14:07:18.569126Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.549193Z","title":"Deepsweep: An evalua- tion framework for mitigating dnn backdoor attacks using data augmentation","venue":null,"work_id":"80fdb147-b919-4bc2-9b88-9f7120641a99","year":2021},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.319354Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:db595997622974574c8e5be4495f0ad3bdc4e142f46a3f1ca09e6338e51006ea","observation_id":"74debac1-6d2f-4efe-904e-7222293fee62","resolution":{"observed_at":"2026-08-12T14:07:18.553331Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2204.06125","last_updated":"2022-04-13T01:10:33Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2022-04-13T01:10:33Z","title":"Hierarchical Text-Conditional Image Generation with CLIP Latents","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2204.06125","snapshot_observed_at":"2026-08-12T14:07:18.324608Z","title":"Hierarchical text-conditional image gen- eration with CLIP latents","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.324608Z"},"links":{"cited_paper":"/paper/2204.06125","citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:82206b1e089a84aab6e0cc918ef61fe16d41f7497db530ba49b3a7b0b836f7f2","observation_id":"2f76c2d9-5233-4d50-8aa7-45b26db67140","resolution":{"observed_at":"2026-08-12T14:07:18.324608Z","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":"2026-08-12T14:07:18.519489Z","title":"How to train your vit? data, augmentation, and regularization in vision transformers","venue":null,"work_id":"18d111d0-2c66-457e-89d9-5da4ce0c0a8a","year":2022},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.334423Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:6e7cb8af24e86ae7d4c4a5a27ca9eca67bb1fafaec0ee8f80eca867505dafacb","observation_id":"de2b0e10-3a8f-4992-b6ae-ef0d3dca7ce6","resolution":{"observed_at":"2026-08-12T14:07:18.525605Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.503110Z","title":"Preserving se- mantic neighborhoods for robust cross-modal retrieval","venue":null,"work_id":"2580dbe7-16fc-4b2e-82fa-72716f7773b0","year":2020},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.338700Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:6483873e80714eaefb2d3debd9daf886340bc966bfbe25271aafb44a27e53d26","observation_id":"1ed620cc-5203-4ce9-b5d2-fe648741b603","resolution":{"observed_at":"2026-08-12T14:07:18.507999Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.488995Z","title":"Data poisoning attacks against federated learning systems","venue":null,"work_id":"b456a111-171d-41a3-b812-746129154ee0","year":2020},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.342973Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:583523abd708d44ff1962dea5cf75a8fcfc1f82bfdfc895b2dd9632659a678c9","observation_id":"392d6bd5-9f0d-4a12-b3e5-7fd779d0948a","resolution":{"observed_at":"2026-08-12T14:07:18.493933Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.473453Z","title":"Bppattack: Stealthy and efficient trojan attacks against deep neural networks via image quantization and contrastive adversar- ial learning","venue":null,"work_id":"0670a945-a5e2-458a-9586-83813fdf6096","year":2022},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.347193Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:c2e3946b98feae1095e4824588f305f8f8f4295e06248d400fcb71e4dc61f90a","observation_id":"145791da-b677-432d-b099-8417b836afec","resolution":{"observed_at":"2026-08-12T14:07:18.478825Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.443060Z","title":"Multimodal fake news detection via clip-guided learning","venue":null,"work_id":"ba9ae76a-1361-4529-8b5d-0a631db27e83","year":2023},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.363722Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:648285c50f80997167b3ff1897e526616189e184c27148ff3baf175a567db50c","observation_id":"f6b293d9-0425-4cfc-8ac6-60be33fdf246","resolution":{"observed_at":"2026-08-12T14:07:18.448083Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.673798Z","title":"Is data cluster- ing in adversarial settings secure? In Proceedings of the 2013 ACM Workshop on Artificial Intelligence and Secu- rity, page 87–98, New York, NY , USA, 2013a","venue":null,"work_id":"da0e055c-dc2a-479f-a7f7-02c033b8953e","year":2013},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":2012,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.279794Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:59997b698b958a2689e8855116843de1dd703caf5e580241da0b684458a67f6b","observation_id":"1ee4be82-6a1b-4305-b664-16740eecdb1e","resolution":{"observed_at":"2026-08-12T14:07:18.678551Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2211.08229","last_updated":"2024-02-29T21:26:45Z","snapshot_observed_at":"2026-08-09T05:21:18.347239Z","submitted_at":"2022-11-15T15:48:28Z","title":"CorruptEncoder: Data Poisoning based Backdoor Attacks to Contrastive Learning","version":5},"cited_work":{"arxiv_id":"2211.08229","doi":null,"metadata_source":"pith","pith_arxiv_id":"2211.08229","snapshot_observed_at":"2026-08-12T14:07:18.405710Z","title":"CorruptEncoder: Data Poisoning based Backdoor Attacks to Contrastive Learning","venue":"cs.CR","work_id":"ca8d4a5d-51b7-4cfd-aee2-b60e8e066c1e","year":2022},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":2014,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.357394Z"},"links":{"cited_paper":"/paper/2211.08229","citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:46be3fe4590837d2b162fc6e44bfc3317ada7cf07eb4c521357453933488a72f","observation_id":"9c57a195-0c2c-4209-8775-b5d7b7efd9e4","resolution":{"observed_at":"2026-08-12T14:07:18.412153Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.459186Z","title":"Robust contrastive language-image pretraining against data poisoning and backdoor attacks","venue":null,"work_id":"a4ebe76c-4755-4978-8e64-041c546ffc5c","year":2023},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":2015,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.352100Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:2311ef2ee62839cfa4a0983126862a8682624e996a8e8fb8aab37dd8368e9068","observation_id":"b120e6a1-ea11-4978-952a-cfd239354c0b","resolution":{"observed_at":"2026-08-12T14:07:18.464267Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.637637Z","title":"Backdoor attack with imperceptible input and latent modification","venue":null,"work_id":"53837c3a-5486-44aa-9e47-f5cd5a8545e9","year":2022},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":2019,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.288274Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:2c13e90ac106a73e89f02aaf1dd19f0b3127c492e753b2a98b71092434af492d","observation_id":"2c4d67cf-eb40-4f2e-8db8-18049c438504","resolution":{"observed_at":"2026-08-12T14:07:18.644195Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.533619Z","title":"Backdoor attacks on self-supervised learning","venue":null,"work_id":"314a5970-57c2-455b-8a8f-d0169d2b1a1c","year":2022},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":2020,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.329719Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:8b0861d5785435f29c792cc54c867c892c98ceb415dfa63fbcac3e35c20160fb","observation_id":"c82d5844-03fb-480b-a3dd-4f5f2bbd7aa9","resolution":{"observed_at":"2026-08-12T14:07:18.538198Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.688059Z","title":"Cleanclip: Mitigating data poisoning attacks in multimodal contrastive learning","venue":null,"work_id":"916f90b7-7fdb-47eb-8e43-f3c181482103","year":2023},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":2021,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.275535Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:6ef2d3a77d00e6f670ecf486d4721569b83372e4d49d1ef25a9b3b14ce87786f","observation_id":"2a4e073c-c873-40a6-bdf8-331c20920284","resolution":{"observed_at":"2026-08-12T14:07:18.692860Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.593653Z","title":"Anti-backdoor learning: Training clean models on poisoned data","venue":null,"work_id":"8054b360-b08a-4980-84de-05537fe026f8","year":2021},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":2022,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.303932Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:d7f55d4d88c17c84e594fbe2b7660c7b89883416425262e70c8206e4b4808e0e","observation_id":"a9017a94-1183-4791-a715-cbe08294a9c2","resolution":{"observed_at":"2026-08-12T14:07:18.597792Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+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":"2026-08-12T14:07:18.656254Z","title":"org/blog/2023-03-30-vicuna, 1(2):3","venue":null,"work_id":"6b3bfe56-dca3-41e9-a047-9dd072e4d0dd","year":2023},"citing_paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment","version":1},"reference_index":2023,"source":"pdf_text","source_observed_at":"2026-08-12T14:07:18.283753Z"},"links":{"citing_paper":"/paper/2411.15673"},"observation_digest":"sha256:faa89c3c8dbcffb019e82c910e2ebfcf8f45ea1ea6bdb0171580bfa367bd2ca6","observation_id":"68a9d7d3-af63-4112-887a-1b7856fe504e","resolution":{"observed_at":"2026-08-12T14:07:18.662806Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-13T06:32:02.005865+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2411.15673","last_updated":"2024-11-23T23:51:52Z","latest_version":1,"primary_category":"cs.CV","snapshot_observed_at":"2026-08-12T13:59:48.714291Z","submitted_at":"2024-11-23T23:51:52Z","title":"Semantic Shield: Defending Vision-Language Models Against Backdooring and Poisoning via Fine-grained Knowledge Alignment"},"reference_resolution":{"displayed":20,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":1,"verified_exact":1,"verified_fuzzy":18},"total_outbound_references":20},"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":"2026-08-13T06:32:02.005865+00:00","source":"crossref"},{"observed_at":"2026-08-13T06:31:53.387327+00:00","source":"retraction_watch"}],"thesis":"As of 13 August 2026, this Paper Citation Record lists 20 of 20 outbound references and 0 inbound Pith citation observations for arXiv:2411.15673."}