{"as_of":"2026-08-08T21:28:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:2df49124e665acc553c6f005c99153c286ff534e4844bc4ade118dbe01bd3212","coverage":[{"denominator":49,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":49,"source":"paper_references, paper_reference_links","source_observed_at":"2026-07-12T01:12:13.747675Z","state":"measured"},{"denominator":49,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":49,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-08T06:32:00.761636+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2607.03612/citation-record","integrity":"/paper/2607.03612/integrity","json":"/paper/2607.03612/citation-record.json","paper":"/paper/2607.03612"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"IJCV120(2), 153–168 (2016)","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:1da5b039726fcbeaa26e2338065e78efaa81c2593fd043b080b55704927416da","observation_id":"02fd5b8f-cef1-425a-abdc-4b6a3f846227","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: Proceedings of the 29th ACM international conference on architectural support for programming languages and operating systems, volume 2","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:0b2acd8ddc6bc29222996ba7b016c75fb2ce76163e5557defd4de9e1b7559195","observation_id":"3ff24134-e0f9-4d8c-b07e-bb862f90c9ab","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2210.09461","last_updated":"2023-03-01T19:45:11Z","snapshot_observed_at":"2026-07-06T14:06:56.291161Z","submitted_at":"2022-10-17T22:23:40Z","title":"Token Merging: Your ViT But Faster","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2210.09461","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2210.09461 (2022)","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2210.09461","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:aebf6429f0b746634f5580d402ac496cd09c000780086e808fc11b54961bc03b","observation_id":"6955a132-fd8d-4a2e-8c16-10ce5e521761","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPRW","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:7b9be96c6f238ef021f4c8a34554fbfdd5e3096441ac773c361bbfe234473d13","observation_id":"911c8dd2-40d8-4838-bd17-738ce95fbeb6","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: ICCV","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:1a9eb23ae57611caf0544dcff8fdff4ea09667f9acffddb5b629f2fae54caae0","observation_id":"d688dc1d-bc96-4ea8-a293-9a37f7bdadd4","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2511.14751","last_updated":"2026-05-14T15:47:56Z","snapshot_observed_at":"2026-08-02T04:22:12.662828Z","submitted_at":"2025-11-18T18:52:22Z","title":"Co-Me: Confidence-Guided Token Merging for Visual Geometric Transformers","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2511.14751","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2511.14751 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2511.14751","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:ad46ed7dc7656ec852050fcd6bf9dfbb201d89aa65876f245c32f6c82a333b58","observation_id":"dc13eab0-183b-4e02-9d6f-cc504c11e80c","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:49f8d6765840e94cd6f1774faf41e872308ad15b018e7c94ff8604ce802ab643","observation_id":"143b16e0-7d88-484e-9f53-08f66ecb95ea","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2309.16588","last_updated":"2024-04-12T09:38:33Z","snapshot_observed_at":"2026-08-07T09:40:32.614733Z","submitted_at":"2023-09-28T16:45:46Z","title":"Vision Transformers Need Registers","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2309.16588","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2309.16588 (2023)","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2309.16588","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:879fee20e8af57ef2bac6b653c1d2804bc6a21f5a3673df47f97e10c3b4eece0","observation_id":"756a207a-1c00-4ed5-80e5-d20fd1cebb66","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2010.11929","last_updated":"2021-06-03T13:08:56Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2020-10-22T17:55:59Z","title":"An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2010.11929","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2010.11929 (2020)","venue":null,"work_id":null,"year":2010},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2010.11929","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:edd9fcd74a628bdf09a67589081a1d2bc25edd27d9c4859a4fe7745834328e85","observation_id":"522207ae-4869-4059-b0d5-f214c1b0b2db","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"Foundations and Trends in Computer Graphics and Vision9(1-2), 1–148 (2015)","venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:886d72419a47259e3d12b40ed3728064fb0b99daa233cac6c664171f2681d98d","observation_id":"bdeabe95-9ce9-447c-82b3-5731eb2cb229","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"NeurIPS37, 52481–52515 (2024)","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:6820f3aebdaca54497be2bcdf7379a3b4202b6eaa23b5bd3e6a6c3aa243c15ab","observation_id":"ff0120f1-a092-4f59-a181-f5b31cc5d7ba","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2509.13414","last_updated":"2026-01-23T18:59:33Z","snapshot_observed_at":"2026-07-06T22:30:07.020487Z","submitted_at":"2025-09-16T18:00:14Z","title":"MapAnything: Universal Feed-Forward Metric 3D Reconstruction","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2509.13414","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2509.13414 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2509.13414","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:bf94d0d236c1b9e84a5218250205c9c3b07fd6f238fade66f4c2bab020b250c5","observation_id":"c8e35f13-f258-4194-81af-081bdecf0a77","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2502.20766","last_updated":"2025-02-28T06:34:53Z","snapshot_observed_at":"2026-08-07T17:40:46.337382Z","submitted_at":"2025-02-28T06:34:53Z","title":"FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2502.20766","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2502.20766 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2502.20766","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:4939b3b4080f6adb51ef10bd2fcea88de412109c7339286f17ce0f5f71009b1e","observation_id":"d930d33b-b247-4e1f-8498-963cfe56059b","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: ECCV","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:aa5b398324dfad6341a26fd6b103e47074ecec804aa211ace37b3296c4f100b9","observation_id":"3925e007-77a8-4e0b-a54c-23faf74aa94c","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2511.10647","last_updated":"2025-11-13T18:59:53Z","snapshot_observed_at":"2026-07-06T22:35:46.018050Z","submitted_at":"2025-11-13T18:59:53Z","title":"Depth Anything 3: Recovering the Visual Space from Any Views","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2511.10647","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2511.10647 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2511.10647","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:ef4630995468ac295b251e7461219a24296b88ad51216f2321392af3c1678e23","observation_id":"55824c7f-7009-4721-b853-58257ea78413","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"IEEE Transactions on Robotics33(5), 1255–1262 (2017)","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:429fbeacaacfd14ef9ae84e587fbea7585b0a28ea6cf154513051e86e885d653","observation_id":"2f2ff326-4a13-4607-a9ba-a97c971c5549","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2304.07193","last_updated":"2024-02-02T10:24:09Z","snapshot_observed_at":"2026-08-06T05:58:29.182448Z","submitted_at":"2023-04-14T15:12:19Z","title":"DINOv2: Learning Robust Visual Features without Supervision","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2304.07193","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2304.07193 (2023) SAF3R 17","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2304.07193","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:b20f793fc0b0c822f6d5574b71ff870ab330994cbac03757ad79f4396606ffb9","observation_id":"4a4a6849-06e9-4dfc-a300-00bf8ff93999","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: ICCV","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:6d561e1fd82c166d5b1aaf619e6efcffcb46d8a131a054b20835c2c9a30cd737","observation_id":"e40b560a-1d50-4402-85eb-d3cff6e2d041","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2603.08055 (2026)","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:41b2fb0b5518f8ae4f94142fa0ea42ef148bb08b0d4fe3f315aeda051fef661e","observation_id":"2e708139-34f5-4b5a-8543-117c0601564a","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:2be975707a1ddbd86881c5406129c81387f3333501487fa664ef978b498d12aa","observation_id":"65d11fb2-c5a3-45bd-8e6e-34c0fa323811","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: ECCV","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:f5357f6593345c8063ff55cca4faec6a5669df69fe0e180490f5c7a91b84c8a4","observation_id":"e440fb5b-fd95-43bb-b669-448c1d7e241e","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:001678d1a9e41b4077bf710927f526235911143878d8981bcb3c20107621b97d","observation_id":"ffe52869-8be2-4221-acf2-b4da04c40068","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2509.02560","last_updated":"2025-11-09T15:12:33Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2025-09-02T17:54:21Z","title":"FastVGGT: Training-Free Acceleration of Visual Geometry Transformer","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2509.02560","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2509.02560 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2509.02560","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:f1a7d369b591c640686929e52d33199abc5c9e233923ad419c16a8db2cfc5f6a","observation_id":"a9398ab6-b91c-4eb4-aff9-f4930eff4e73","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2013},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:28b460e543ce1e7d5db39b029103d607d3118d6a6b8a9ef94a691a07d0d7e680","observation_id":"64c3498a-cb99-42b7-bc4c-6e1c3407ece6","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2512.04939 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:9abb2ae2b7ddfffd17289da2fc0c6d9a03d220360ae0bbac0e47c266e26f58fa","observation_id":"2899d471-f57e-4eb8-b9d1-f05f9f5a7e4d","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2512.02541 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:41ecf7f6e56e12624e10280c969ec478e56146e6660ea61b4f43e8acd3c9b3c3","observation_id":"c9d4cce6-b84d-466d-92f2-db0fcd87a991","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"The International Journal of Robotics Research 45(6), 839–857 (2026)","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:ea91b006733091f63bc8de75ce00f168bbf863fdd373cd3e1e0df924370345f1","observation_id":"203d41c0-f78a-4c7d-815a-cfa5a1d57702","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: Proceedings of the 3rd ACM SIGPLAN International Workshop on Machine Learning and Programming Languages","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:15eab900c18769b696e39d2ff40906bfaee291958c42eeeb12863e2b6f6f6cce","observation_id":"bcf49402-4ce3-4d37-8a39-49ab0905192e","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"NeurIPS30(2017)","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:2fea761afbc3b20d9aa7ce57675331e6acb5df212a9d1dfb79335dd6b70a598f","observation_id":"b8812dd7-2e31-4e3f-b7c1-f9c9977ec3be","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2509.07120","last_updated":"2026-05-20T17:17:47Z","snapshot_observed_at":"2026-07-06T22:26:08.072526Z","submitted_at":"2025-09-08T18:16:09Z","title":"Block-Sparse Global Attention for Efficient Multi-View Geometry Transformers","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2509.07120","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2509.07120 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2509.07120","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:253d20f1f0eba92e0e66dd720163ebbf9fe09d35238d68ce3cf0295549951c99","observation_id":"41f366c2-01a0-4fc0-aac2-735efaa20433","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:a31c5c2353618f9f40214c96a7cbb53100dee58f560c1e8a8d61464156dab522","observation_id":"8b1ad4af-7186-4978-a8b7-e663cf9d797c","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:7999246b4117c3c1f1a7a49e092b65b0c6f5dc5a9d72a121f5b5757a0bd494b9","observation_id":"db887a9c-2076-4863-a0ca-a474f580bb8e","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:632229a4a10ba9f1d21a379b99f943877acfdc0b8098b5bf16c837bcfc33c38e","observation_id":"cf009a61-41fd-4788-a1fb-a3057ffef078","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:d0bea7ec53a9206984387d4545042d74c696dc50d67942e11e5112166667e2b5","observation_id":"10cac987-81bc-4e55-a841-7794f4e9fff3","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2511.21317 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:bff48a29ade337c9d1331e08ffeb4f9630b52d74104164f88af6c76233dc3943","observation_id":"21442a05-a030-4346-8c77-31fdd14039ec","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2507.13347","last_updated":"2026-03-07T07:01:59Z","snapshot_observed_at":"2026-08-02T12:12:15.465550Z","submitted_at":"2025-07-17T17:59:53Z","title":"$\\pi^3$: Permutation-Equivariant Visual Geometry Learning","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2507.13347","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2507.13347 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2507.13347","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:9f1de1b445e8d9c0e796c590dda2dcaa339fd7e5ece7ba82b3f2f6562a61871a","observation_id":"9b4c4ccb-3df7-467a-8596-0cfd688af981","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2512.01540 (2025) 18 J","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:f7f03e6ad1051f71411af251746dd9d629e9393176eddadbe6173cae7f7cb79c","observation_id":"bada8be5-007d-43bb-ac1b-8e162a7fe16d","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2502.01776","last_updated":"2025-04-27T00:10:11Z","snapshot_observed_at":"2026-08-06T01:54:01.546928Z","submitted_at":"2025-02-03T19:29:16Z","title":"Sparse VideoGen: Accelerating Video Diffusion Transformers with Spatial-Temporal Sparsity","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2502.01776","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2502.01776 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2502.01776","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:7309a9e7a5c0cc35f0be46b7e1eccc4d271cf127f1e8f5414097e547d0ff828e","observation_id":"7acff67a-d000-43a3-9681-d355d9067ea5","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2410.10819","last_updated":"2024-10-14T17:59:58Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-10-14T17:59:58Z","title":"DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2410.10819","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2410.10819 (2024)","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2410.10819","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:73650ebcf1ea9de1d335e0e6c434278e93935e252ae0cc4aa01d009aed28814b","observation_id":"8d15d824-9795-4610-bacc-1f854efb583d","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2309.17453","last_updated":"2024-04-07T00:56:53Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-09-29T17:59:56Z","title":"Efficient Streaming Language Models with Attention Sinks","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2309.17453","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2309.17453 (2023)","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2309.17453","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:77ca87430fadad0b43fc1c0f1ab6f36c332cffd94bd5d04ab2853eea70b58777","observation_id":"6541b661-8542-4a0b-821d-735112a1eff0","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: ICML (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:4e854e382b7afd530f332f193bfd5ab5f06e51d11f1528e032845b863c5237d2","observation_id":"abc168a6-533b-48ab-af20-94a30764398e","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:c05e25f3abbe0b2c8f6a2eb2454e404343c25c686e3e816d46e3e3c9d8e57048","observation_id":"d8270b8a-edce-4f50-bbbd-b8cbb24d5480","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: CVPR","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:0cf70a3db5dac7ad4c3f79e4b2ee72800a5e881594133e58fe3b54ebb59faf5d","observation_id":"23c1a156-4207-407c-a49c-133ffa41cff6","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.18875","last_updated":"2026-05-06T23:08:29Z","snapshot_observed_at":"2026-08-02T05:41:40.015287Z","submitted_at":"2025-05-24T21:30:29Z","title":"Sparse VideoGen2: Accelerate Video Generation with Sparse Attention via Semantic-Aware Permutation","version":5},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.18875","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2505.18875 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2505.18875","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:704a53a5a22e1a251f68ccfa5e77aba59f179880cd2898534e82402af278417f","observation_id":"7a9a6fca-032f-47d2-8b0b-6774ee529a24","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: ICCV","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:00a92021a3381f1cd9b00f4795021d36126b5a4aea2bfe57c0757d071bae4352","observation_id":"9c8421e4-e9c2-49a1-a50d-d73bf9b67deb","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:ae0d6065cb8c9b9ccd4a7f94645fc3f03fc44ca07fe79f44a0a77985ffd229d3","observation_id":"54a21151-f9ee-42dd-89f4-f70280ec5d1b","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","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":"2026-07-12T01:12:13.747675Z","title":"In: ICML (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:92796f0aa2ebf35b06776afe7550463448244e2852c53ad984d3a2ff0f4f40ae","observation_id":"d8027bcf-8d48-40c6-9ddd-047aa0178959","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2502.04507","last_updated":"2025-06-04T23:21:39Z","snapshot_observed_at":"2026-08-05T18:16:36.053313Z","submitted_at":"2025-02-06T21:17:09Z","title":"Fast Video Generation with Sliding Tile Attention","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2502.04507","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"arXiv preprint arXiv:2502.04507 (2025)","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/2502.04507","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:e51e7e0cbeb6402f4afc58308b1faff00d074fecc14ae8e3fd7ec28f13280a0f","observation_id":"969de086-4a8e-42c1-962e-027bb1f1ff16","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1805.09817","last_updated":"2018-05-24T17:58:02Z","snapshot_observed_at":"2026-07-06T06:41:05.545426Z","submitted_at":"2018-05-24T17:58:02Z","title":"Stereo Magnification: Learning View Synthesis using Multiplane Images","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1805.09817","snapshot_observed_at":"2026-07-12T01:12:13.747675Z","title":"9:Representative global attention patterns across layers (G) and heads (H) in four F3R transformers on the7Scenes Chessscene","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-07-12T01:12:13.747675Z"},"links":{"cited_paper":"/paper/1805.09817","citing_paper":"/paper/2607.03612"},"observation_digest":"sha256:231eac18f7800d64574ecf536fdc7dded494c209622dba62c25d37e95a38f9fc","observation_id":"6a50ff5a-4556-4269-9b79-937159d1aa50","resolution":{"observed_at":"2026-07-12T01:12:13.747675Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2607.03612","last_updated":"2026-07-03T22:01:35Z","latest_version":1,"primary_category":"cs.CV","snapshot_observed_at":"2026-08-03T07:17:39.441528Z","submitted_at":"2026-07-03T22:01:35Z","title":"SAF3R: Dynamic Sparse Attention for Feed-Forward 3D Reconstruction Transformers"},"reference_resolution":{"displayed":49,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":49,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":49},"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":"2026-08-08T06:32:00.761636+00:00","source":"crossref"},{"observed_at":"2026-08-08T06:31:55.24221+00:00","source":"retraction_watch"}],"thesis":"As of 8 August 2026, this Paper Citation Record lists 49 of 49 outbound references and 0 inbound Pith citation observations for arXiv:2607.03612."}