{"as_of":"2026-08-09T00:53:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:960cbfed43a7340ffd41c42fcdde0a605ae900b64714220adf5a47ffa2a0f988","coverage":[{"denominator":91,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":91,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-03T19:24:39.437108Z","state":"measured"},{"denominator":92,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":92,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-08T06:32:00.761636+00:00","state":"measured"},{"denominator":1,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":1,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-03T07:38:50.237882Z","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":[{"citation":{"cited_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2512.00814","snapshot_observed_at":"2026-08-03T07:38:50.237882Z","title":"Irpo: Boosting image restoration via post-training grpo,","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2601.19506","last_updated":"2026-06-30T07:23:59Z","snapshot_observed_at":"2026-08-08T08:13:54.446239Z","submitted_at":"2026-01-27T11:50:31Z","title":"Bridging Information Asymmetry: A Hierarchical Framework for Deterministic Blind Face Restoration","version":4},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-03T07:38:50.237882Z"},"links":{"cited_paper":"/paper/2512.00814","citing_paper":"/paper/2601.19506"},"observation_digest":"sha256:0779704a437a9f02bb8327adc95ef4c60479be589c3a1cabebec526baeb96811","observation_id":"8a52bf5e-2f8e-46dd-967e-9ed89c3bb46c","resolution":{"observed_at":"2026-08-03T07:38:50.237882Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"links":{"evidence":"/evidence","html":"/paper/2512.00814/citation-record","integrity":"/paper/2512.00814/integrity","json":"/paper/2512.00814/citation-record.json","paper":"/paper/2512.00814"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-03T19:24:37.996899Z","title":"A high-quality denoising dataset for smartphone cameras","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:37.996899Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:8111dc5daabd50673fbe7988ecf9991afb8f3f92eb798716b40c95a6087c298c","observation_id":"2f23b101-2dd6-44a4-9f67-696da170171f","resolution":{"observed_at":"2026-08-03T19:24:37.996899Z","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":"2026-08-03T19:24:38.059864Z","title":"Contour detection and hierarchical image seg- mentation.TPAMI, 2010","venue":null,"work_id":null,"year":2010},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.059864Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:a003158a8155c098cb4a5cda0d9c04c7b8d8ab2d685e2ef3374662149499392a","observation_id":"5572575a-6c74-4a1a-a653-a01ff33033a0","resolution":{"observed_at":"2026-08-03T19:24:38.059864Z","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":"2026-08-03T19:24:38.156204Z","title":"Not just streaks: Towards ground truth for single image derain- ing","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.156204Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:b1cbcc5642ed8d72194067f356cbca6f3836e8feb6cb7f8fbadce17cd853e1f9","observation_id":"4f233de6-968d-4402-ad41-8892a8c76723","resolution":{"observed_at":"2026-08-03T19:24:38.156204Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2204.05862","last_updated":"2022-04-12T15:02:38Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2022-04-12T15:02:38Z","title":"Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2204.05862","snapshot_observed_at":"2026-08-03T19:24:38.293457Z","title":"Training a helpful and harmless assistant with reinforcement learning from human feedback.arXiv preprint arXiv:2204.05862, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.293457Z"},"links":{"cited_paper":"/paper/2204.05862","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:dfae915992a74cc348ab44aa62b7a07beebc8e9f959ce3c045a4df0d278a437f","observation_id":"4909783c-d742-4d1a-9039-1b016d4a8d6a","resolution":{"observed_at":"2026-08-03T19:24:38.293457Z","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":"2026-08-03T19:24:38.449550Z","title":"The perception-distortion tradeoff","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.449550Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:11107a125ba27187f81ad890cc81ec42cb423f231b7e9f421353a8cdf16ccd09","observation_id":"cfde42bf-4956-4698-a6dd-312df9c104e6","resolution":{"observed_at":"2026-08-03T19:24:38.449550Z","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":"2026-08-03T19:24:38.534541Z","title":"Dehazenet: An end-to-end system for single image haze removal.TIP, 2016","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.534541Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:cf04ece7685eafbd1b0223160b91f030859e0a35d1392ec45c7837c6786943e7","observation_id":"6afe147d-bb5f-40e1-85b3-2d0b586dfc0d","resolution":{"observed_at":"2026-08-03T19:24:38.534541Z","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":"2026-08-03T19:24:38.614814Z","title":"Hinet: Half instance normalization network for image restoration","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.614814Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:d6505aa718f25a71e1ee77ab717eb7e080924d3dc5871c5b3aa66f2bfc3a1cb6","observation_id":"feea8c42-a54f-4d9d-b45d-dba2ffec2436","resolution":{"observed_at":"2026-08-03T19:24:38.614814Z","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":"2026-08-03T19:24:38.669070Z","title":"Simple baselines for image restoration","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.669070Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:379054b925d1b02f0905808172807a92769b3fe6695c55bcc574a45268643cff","observation_id":"c39cf595-412a-450a-b4ef-a7ae6e64d275","resolution":{"observed_at":"2026-08-03T19:24:38.669070Z","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":"2026-08-03T19:24:38.751918Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.751918Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:8a5c0ad10d301c58664ecdf8b0dd39fa9f7ddb18b8fecf0bf4b11fe778e97e21","observation_id":"ae647bd9-a64c-4318-b33c-9a7a4ae6c384","resolution":{"observed_at":"2026-08-03T19:24:38.751918Z","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":"2026-08-03T19:24:38.829784Z","title":"AdaIR: Adap- tive all-in-one image restoration via frequency mining and modulation","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.829784Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:83da7ffbfbe2c498e55073f38de27878d8997cef8b0587cfc7372c745ec478d9","observation_id":"60736161-6a49-4168-bee6-5fa49cb1ebf8","resolution":{"observed_at":"2026-08-03T19:24:38.829784Z","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":"2026-08-03T19:24:38.912046Z","title":"Color image denoising via sparse 3d col- laborative filtering with grouping constraint in luminance- chrominance space","venue":null,"work_id":null,"year":2007},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.912046Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:7e3f134dc78a156d60caeb3972865adb223c6d59a1131ef00f93165ac3583464","observation_id":"217771f7-d235-42b4-aa78-ef9fd1cc1374","resolution":{"observed_at":"2026-08-03T19:24:38.912046Z","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":"2026-08-03T19:24:38.966234Z","title":"Image super-resolution using deep convolutional net- works.IEEE transactions on pattern analysis and machine intelligence, 38(2):295–307, 2015","venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:38.966234Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:59387acc91dd51fd7b8320f68e7ee457ab078c3a5a66eacfa6744478d72c906d","observation_id":"b221ce87-db04-455e-a6b3-0304171e78d3","resolution":{"observed_at":"2026-08-03T19:24:38.966234Z","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":"2026-08-03T19:24:39.046181Z","title":"Fd-gan: Generative adversarial networks with fusion- discriminator for single image dehazing","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.046181Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:b9a50496b9f56df2c6e432be9f5f5d64113f9a6aee100bb99cf1452efc5b6032","observation_id":"fec16136-fe76-4d5e-8920-29cc6f8fed5b","resolution":{"observed_at":"2026-08-03T19:24:39.046181Z","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":"2026-08-03T19:24:39.101365Z","title":"A general decoupled learn- ing framework for parameterized image operators.TPAMI,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.101365Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:af596edde6e5be328e046525c22fd5c234f9d53284c2c2f075722392c212b536","observation_id":"99516f45-064b-491c-8435-65b856e1e48c","resolution":{"observed_at":"2026-08-03T19:24:39.101365Z","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":"2026-08-03T19:24:39.155783Z","title":"Dy- namic scene deblurring with parameter selective sharing and nested skip connections","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.155783Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:6ae541b5a7af542293f1dd08cc06e488caafb9078d1637cb2f139213c459154b","observation_id":"d857fd32-c8c1-482b-b4ca-cb11b79c0bbf","resolution":{"observed_at":"2026-08-03T19:24:39.155783Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2411.11504","last_updated":"2024-11-18T12:04:52Z","snapshot_observed_at":"2026-07-06T19:51:54.660531Z","submitted_at":"2024-11-18T12:04:52Z","title":"Search, Verify and Feedback: Towards Next Generation Post-training Paradigm of Foundation Models via Verifier Engineering","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2411.11504","snapshot_observed_at":"2026-08-03T19:24:39.207564Z","title":"Search, verify and feedback: Towards next generation post- training paradigm of foundation models via verifier engineer- ing.arXiv preprint arXiv:2411.11504, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.207564Z"},"links":{"cited_paper":"/paper/2411.11504","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:2fa7c7ae583daaac5e1b6cc534e48742470ddede365de1fe875cf24ff067832d","observation_id":"ed9af779-95b3-4d65-91f3-89ad7edd0086","resolution":{"observed_at":"2026-08-03T19:24:39.207564Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.12948","last_updated":"2026-01-04T03:57:36Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2025-01-22T15:19:35Z","title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.12948","snapshot_observed_at":"2026-08-03T19:24:39.232544Z","title":"Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning.arXiv preprint arXiv:2501.12948, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.232544Z"},"links":{"cited_paper":"/paper/2501.12948","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:bfd315374f0b35c7dbbddaa2bb8c08c8313d46cb3ac9ec78ab915acd0cea43e2","observation_id":"9dfee3a1-7e99-44c9-afb0-998f9e7469f7","resolution":{"observed_at":"2026-08-03T19:24:39.232544Z","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":"2026-08-03T19:24:39.235720Z","title":"Toward convolutional blind denoising of real pho- tographs","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.235720Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:fc4cd4b3a2736b1a8ff8daad42a7c81f52713a260e85487bbce82a199cfe931e","observation_id":"3eeb0094-159c-4336-8656-eea6fdcf3958","resolution":{"observed_at":"2026-08-03T19:24:39.235720Z","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":"2026-08-03T19:24:39.238981Z","title":"Face super-resolution guided by 3d facial priors","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.238981Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:0c7d08de1aa7b68cae8fb4687368815e4050b7b687836e0ba7d8174f4f367e2f","observation_id":"23512d5f-23e2-4032-9e38-ae17b5a7cdab","resolution":{"observed_at":"2026-08-03T19:24:39.238981Z","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":"2026-08-03T19:24:39.242111Z","title":"Face restoration via plug-and-play 3d facial priors.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(12):8910–8926, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.242111Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:856bcb81f81a4c2449113f75d8dccdba2f39eb40cda30116944f3159f77434fc","observation_id":"6628b36f-500b-4fe5-be82-172807e1b07b","resolution":{"observed_at":"2026-08-03T19:24:39.242111Z","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":"2026-08-03T19:24:39.244753Z","title":"Single image super-resolution from transformed self-exemplars","venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.244753Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:5282a25c1dda4378a56c984a207244f12c2f90771042b4d90f9dffc681a207a3","observation_id":"571474a7-a983-48b8-8226-3240e007d558","resolution":{"observed_at":"2026-08-03T19:24:39.244753Z","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":"2026-08-03T19:24:39.247354Z","title":"Hunyuan3d-omni: A unified framework for controllable generation of 3d assets","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.247354Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:861ecbf3e3eb766623cfc2c728cf754873c1cbd86703423067b73e976b828acb","observation_id":"90e2b5bb-80df-4329-970f-2f141a78e528","resolution":{"observed_at":"2026-08-03T19:24:39.247354Z","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":"2026-08-03T19:24:39.250023Z","title":"Loli-street: Bench- marking low-light image enhancement and beyond","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.250023Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:3313b5c2683fea8e4eae6c498d9b5d103bb057068850d90844a3a385467eb42e","observation_id":"ae444e51-b47e-4ecc-99a2-84262b80f6ae","resolution":{"observed_at":"2026-08-03T19:24:39.250023Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2412.16720","last_updated":"2026-04-30T02:46:40Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2024-12-21T18:04:31Z","title":"OpenAI o1 System Card","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2412.16720","snapshot_observed_at":"2026-08-03T19:24:39.252897Z","title":"Openai o1 system card.arXiv preprint arXiv:2412.16720, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.252897Z"},"links":{"cited_paper":"/paper/2412.16720","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:5b29f7cecfa0539dd0c0077488bdd83488b6c71eed90d494bd3bde3287b3b05c","observation_id":"cb1880f5-8d44-4726-841d-20291efb4537","resolution":{"observed_at":"2026-08-03T19:24:39.252897Z","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":"2026-08-03T19:24:39.255708Z","title":"Supervised learning of image restoration with convolutional networks","venue":null,"work_id":null,"year":2007},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.255708Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:6a989bfa50f00a253cb2faa2c3919e77810e5f81ed990deab267296093450463","observation_id":"16898427-0028-4a60-a018-f88e393b0630","resolution":{"observed_at":"2026-08-03T19:24:39.255708Z","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":"2026-08-03T19:24:39.258211Z","title":"Sonic: Shifting focus to global au- dio perception in portrait animation","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.258211Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:7a0640ba3900f3f12114b7ce50fe2e9b2ad87195804cd34d4d713ce0a291daba","observation_id":"5a70253b-6311-4c84-8cad-972c2e1161fc","resolution":{"observed_at":"2026-08-03T19:24:39.258211Z","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":"2026-08-03T19:24:39.260905Z","title":"Multi-scale progressive fusion network for single image deraining","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.260905Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:a453b86d9d4330fa0d6cdc01fa669d249c52317340ce8f515a149f124cda5a3e","observation_id":"a36b0dca-643d-48b3-ac50-fd4d50436f58","resolution":{"observed_at":"2026-08-03T19:24:39.260905Z","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":"2026-08-03T19:24:39.263414Z","title":"Perceptual losses for real-time style transfer and super-resolution","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.263414Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:1d7c22c7ebdd42907ef5ba3f0e964b5debc6ff2c2a65a830e75bf4b7b0cc4ff9","observation_id":"a62abafa-08c6-478a-98e6-a861b7db506c","resolution":{"observed_at":"2026-08-03T19:24:39.263414Z","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":"2026-08-03T19:24:39.265991Z","title":"Can grpo boost complex mul- timodal table understanding? InProceedings of the 2025 Conference on Empirical Methods in Natural Language Pro- cessing, pages 12642–12655, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.265991Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:8f884380263cf816d1e721f969d67b065b7844759e70af7b09b6748000b7aefa","observation_id":"5815d438-b195-45de-be6e-c6442623e4eb","resolution":{"observed_at":"2026-08-03T19:24:39.265991Z","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":"2026-08-03T19:24:39.268584Z","title":"A survey of post-training scaling in large language models","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.268584Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:d425fe86a17e11364a4662ea8e19bbf2ae37761823fa7c9c6e22501841f30fbe","observation_id":"9bb1866f-e285-4f85-958c-e1db0a10bb99","resolution":{"observed_at":"2026-08-03T19:24:39.268584Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1803.04189","last_updated":"2018-10-29T10:29:23Z","snapshot_observed_at":"2026-07-06T06:27:46.580255Z","submitted_at":"2018-03-12T11:07:58Z","title":"Noise2Noise: Learning Image Restoration without Clean Data","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1803.04189","snapshot_observed_at":"2026-08-03T19:24:39.271347Z","title":"Noise2noise: Learning image restoration without clean data","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.271347Z"},"links":{"cited_paper":"/paper/1803.04189","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:279e869b7a2176ad4d844cbc754bdc757ce44890c7d6dbf81ba212a4a59b9a48","observation_id":"46ff086a-b29b-451d-a5ed-e5b5277cf4b6","resolution":{"observed_at":"2026-08-03T19:24:39.271347Z","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":"2026-08-03T19:24:39.274222Z","title":"Benchmarking single- image dehazing and beyond.TIP, 2018","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.274222Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:8919d149e18b5cb8135f43a81305534945fe9ff28ffe10938541e950ed430b5a","observation_id":"b4e234d8-651b-4a7e-b358-23e1cf831477","resolution":{"observed_at":"2026-08-03T19:24:39.274222Z","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":"2026-08-03T19:24:39.276810Z","title":"All-in-one image restoration for unknown cor- ruption","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.276810Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:39cf80bb62a9c9e9f5be3452ad6c74b8b20cfc558212e1002f16fd616c49840e","observation_id":"4d08bdde-df4d-4f7a-9751-c30c4f09ab02","resolution":{"observed_at":"2026-08-03T19:24:39.276810Z","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":"2026-08-03T19:24:39.279394Z","title":"Real-world deep local motion deblur- ring","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.279394Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:93f49a24e2432fc6b0e5282a41c6c41fb556ae3627966cb00dad1605bb68d3e4","observation_id":"44028547-7c83-42c6-8041-f21b7df47282","resolution":{"observed_at":"2026-08-03T19:24:39.279394Z","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":"2026-08-03T19:24:39.282074Z","title":"T2v- turbo-v2: Enhancing video generation model post-training through data, reward, and conditional guidance design.arXiv preprint arXiv:2410.05677, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.282074Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:f24df2a5340682161642985f837485962512b5242d3df79dc9ae564f8df91760","observation_id":"22843845-0f3a-4030-a75e-976f4f12f2fc","resolution":{"observed_at":"2026-08-03T19:24:39.282074Z","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":"2026-08-03T19:24:39.284556Z","title":"Swinir: Image restoration us- ing swin transformer","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.284556Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:3ad949373bdd9095a58cf81c2fd0f72580ab4efb25ad24c4d071e41e3dfdbbc5","observation_id":"c0c9923c-84e9-421a-a032-08fe25c9cb2d","resolution":{"observed_at":"2026-08-03T19:24:39.284556Z","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":"2026-08-03T19:24:39.287635Z","title":"Vrt: A video restoration transformer.IEEE Transactions on Image Processing, 33:2171–2182, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.287635Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:09b9792815f48a338fa066b4c0b4414e2a6e68a8627d18b74f53e57644d1c0e4","observation_id":"7b729df0-dea9-42fc-bb77-8f0eaefc776e","resolution":{"observed_at":"2026-08-03T19:24:39.287635Z","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":"2026-08-03T19:24:39.290291Z","title":"Vton-handfit: Virtual try-on for arbi- trary hand pose guided by hand priors embedding","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.290291Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:9ed3b2cd20f18fe7d244ced4803233c26bead234cc7569d35d3fe5dde3e8ce81","observation_id":"76c252c4-1828-4503-9b0b-4367180f5ea7","resolution":{"observed_at":"2026-08-03T19:24:39.290291Z","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":"2026-08-03T19:24:39.292767Z","title":"Diffusion adversarial post-training for one-step video generation.arXiv preprint arXiv:2501.08316,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.292767Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:b53facb2e848192556d5354d8a5945b829b76732ced9d3acc3262522f19fba34","observation_id":"fb03eca8-6161-40b6-be92-914c94e237e9","resolution":{"observed_at":"2026-08-03T19:24:39.292767Z","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":"2026-08-03T19:24:39.295640Z","title":"Autoregressive adversarial post-training for real-time inter- active video generation.arXiv preprint arXiv:2506.09350,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.295640Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:faaece57c28b6744f7bbccb53eb3d77d1c87cfd63d7f97c86048fbea4c905602","observation_id":"7feb49c2-ddaf-42fc-be7d-640542e53845","resolution":{"observed_at":"2026-08-03T19:24:39.295640Z","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":"2026-08-03T19:24:39.298091Z","title":"Tape: Task-agnostic prior embedding for image restoration","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.298091Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:d94488428189ea7ec671bb412f1070ba0999de3b30edc758c3e8113dc8d5d3b6","observation_id":"92f3335f-7d4a-421b-a7e6-c36c43c4a119","resolution":{"observed_at":"2026-08-03T19:24:39.298091Z","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":"2026-08-03T19:24:39.300641Z","title":"Moa-vr: A mixture- of-agents system towards all-in-one video restoration.arXiv preprint arXiv:2510.08508, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.300641Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:26bd142720a96de8c53346094ee21f04955895838b5c03ec1780064b003571c7","observation_id":"98c906bd-3048-4648-b010-9c03afa9d842","resolution":{"observed_at":"2026-08-03T19:24:39.300641Z","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":"2026-08-03T19:24:39.303633Z","title":"Two-stage mamba-based diffusion model for image restora- tion.Scientific Reports, 15(1):22265, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.303633Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:2f28f6cc41fe17577e2cab98ab198edae6e33b5be9db35a57eb7d3d25d99ade6","observation_id":"4f1a54a6-0afd-4e23-a383-795350605cf3","resolution":{"observed_at":"2026-08-03T19:24:39.303633Z","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":"2026-08-03T19:24:39.306125Z","title":"Post-training quantization for vision trans- former.Advances in Neural Information Processing Systems, 34:28092–28103, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.306125Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:ebea7217422394ba1670a45d7dc4f720a0d10abc0d32b12334e5cb6a662a45fc","observation_id":"7922e5b3-4e0b-4298-8e41-375e18aa2f7a","resolution":{"observed_at":"2026-08-03T19:24:39.306125Z","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":"2026-08-03T19:24:39.308906Z","title":"Waterloo ex- ploration database: New challenges for image quality assess- ment models.TIP, 2016","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.308906Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:658b872002372b9cf66e070aae9e2a60667b9d93f1616837b3da4878c6a6f15d","observation_id":"768c9bc7-ca53-4a09-96ee-643fdb8f9fcc","resolution":{"observed_at":"2026-08-03T19:24:39.308906Z","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":"2026-08-03T19:24:39.311376Z","title":"A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics","venue":null,"work_id":null,"year":2001},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.311376Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:9d60c99c3f95c5d1ae2b09510a30128165e450a65af3a1575b1b68432c3c8209","observation_id":"b651bc9a-9266-4d96-ab0a-6cc503aae921","resolution":{"observed_at":"2026-08-03T19:24:39.311376Z","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":"2026-08-03T19:24:39.313929Z","title":"Deep generalized unfolding networks for image restoration","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.313929Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:32ca74b10e97c4642c839bcb866eac4a071eecfd25d18af0ae639257e6d207d1","observation_id":"751f7cf0-e40e-46cf-9a69-09c384d2c183","resolution":{"observed_at":"2026-08-03T19:24:39.313929Z","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":"2026-08-03T19:24:39.316536Z","title":"Deep multi-scale convolutional neural network for dynamic scene deblurring","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.316536Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:8a7dfd57898ca7dca2fe80fe2566baec3b4711b80b33f63d8c787f6d5a44bc16","observation_id":"5fcf4bf7-687c-407e-b5dd-8a1ee2e9aa33","resolution":{"observed_at":"2026-08-03T19:24:39.316536Z","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":"2026-08-03T19:24:39.319042Z","title":"Promptir: Prompting for all-in- one image restoration.NeurIPS, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.319042Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:3777da2554d03856be3676015d61fb8ea3433691c7670f46e297a9e6bf98eaee","observation_id":"1a09e818-be8b-4774-a3d5-71633af190db","resolution":{"observed_at":"2026-08-03T19:24:39.319042Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2506.17221","last_updated":"2025-06-25T06:03:22Z","snapshot_observed_at":"2026-08-06T23:28:19.444744Z","submitted_at":"2025-06-20T17:59:59Z","title":"VLN-R1: Vision-Language Navigation via Reinforcement Fine-Tuning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2506.17221","snapshot_observed_at":"2026-08-03T19:24:39.321499Z","title":"Vln-r1: Vision-language navigation via reinforcement fine-tuning.arXiv preprint arXiv:2506.17221,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.321499Z"},"links":{"cited_paper":"/paper/2506.17221","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:810ffb9ddfce7691a2d1d6d65ecacba43bd39153d810edd266fe6b1ce1396dc1","observation_id":"8f6d4493-0bc1-4a99-9994-6333f0c359cd","resolution":{"observed_at":"2026-08-03T19:24:39.321499Z","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":"2026-08-03T19:24:39.324672Z","title":"En- hanced pix2pix dehazing network","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.324672Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:57067b91155c528fa62708418397688d820590629c8c7a5d5044bd18a827c302","observation_id":"f9e9a19b-17ea-4cec-b9b0-61dc61d4d9a3","resolution":{"observed_at":"2026-08-03T19:24:39.324672Z","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":"2026-08-03T19:24:39.327253Z","title":"Learning transferable visual models from natural language supervi- sion","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.327253Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:822e235f8ccc4a9afe552903be94f6284f1bc83299c8f96002bfd3189313d22b","observation_id":"9c7e288e-70c7-436b-ba8e-7c1cea19f9e4","resolution":{"observed_at":"2026-08-03T19:24:39.327253Z","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":"2026-08-03T19:24:39.329871Z","title":"Direct preference optimization: Your language model is secretly a reward model.Advances in neural information processing systems, 36:53728–53741, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.329871Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:df6b6c96882beba6151cfa3b91264d088a311b7eaef9510536de02932e664a77","observation_id":"219231a8-be9c-4b23-ba40-a164ca47320d","resolution":{"observed_at":"2026-08-03T19:24:39.329871Z","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":"2026-08-03T19:24:39.332520Z","title":"Single image dehazing via multi- scale convolutional neural networks","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.332520Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:2b531c21d721d58410e632c1eebb5aadbd5c060c481b183c98c74c78f5c7b876","observation_id":"6fa7d36c-f232-4ca4-aaeb-da0b8d1bcf31","resolution":{"observed_at":"2026-08-03T19:24:39.332520Z","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":"2026-08-03T19:24:39.335265Z","title":"Laion-5b: An open large-scale dataset for training next generation image-text models.Advances in neural in- formation processing systems, 35:25278–25294, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.335265Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:dc30f6186e1bbed34d48d9fa2353831369cc5c51f73a2b0a5d5489f676b930a2","observation_id":"2323ba71-286c-4e37-95c4-b54b8f75c630","resolution":{"observed_at":"2026-08-03T19:24:39.335265Z","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":"2026-08-03T19:24:39.337911Z","title":"Post-training quantization on diffusion models","venue":null,"work_id":null,"year":1972},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.337911Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:d2ee5c456b0ae192e9e9fe0a471de25b725785f6b2c88f20d1ef0a8972d9d67f","observation_id":"747a5489-7f3c-4c55-8d00-5d6628b7e17c","resolution":{"observed_at":"2026-08-03T19:24:39.337911Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.03300","last_updated":"2024-04-27T15:25:53Z","snapshot_observed_at":"2026-08-06T14:58:42.911363Z","submitted_at":"2024-02-05T18:55:32Z","title":"DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.03300","snapshot_observed_at":"2026-08-03T19:24:39.340677Z","title":"Deepseekmath: Pushing the limits of math- ematical reasoning in open language models.arXiv preprint arXiv:2402.03300, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.340677Z"},"links":{"cited_paper":"/paper/2402.03300","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:08ad7868cc6d88c0d0a7da35b10eaaf8a6280ef2eec99cd145e2c243728095c4","observation_id":"62eb4948-55bc-4d9a-99e0-de4e55185b90","resolution":{"observed_at":"2026-08-03T19:24:39.340677Z","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":"2026-08-03T19:24:39.343943Z","title":"Fine-grained image quality assessment for per- ceptual image restoration.arXiv preprint arXiv:2508.14475,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":58,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.343943Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:12194be8bb9ecc8bd8cabedb0a7cda9440329067c4a586d33675bab949282014","observation_id":"f3914dc9-f482-44e9-96b5-5b231900b975","resolution":{"observed_at":"2026-08-03T19:24:39.343943Z","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":"2026-08-03T19:24:39.346709Z","title":"Image de- noising using deep cnn with batch renormalization.Neural Networks, 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":59,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.346709Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:8dd5d6792d74eef681ee7088c2bb429324d19d8f4990897dfccfdbc7ce274e9a","observation_id":"81d81336-1112-4d5d-b41a-27eaec3a0852","resolution":{"observed_at":"2026-08-03T19:24:39.346709Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2302.13971","last_updated":"2023-02-27T17:11:15Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-02-27T17:11:15Z","title":"LLaMA: Open and Efficient Foundation Language Models","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2302.13971","snapshot_observed_at":"2026-08-03T19:24:39.349405Z","title":"Llama: Open and efficient foundation language models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":60,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.349405Z"},"links":{"cited_paper":"/paper/2302.13971","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:5ca818570a1085a3cc8708802d398584d02f71861fbf15f9ebee4fee3e6feb4a","observation_id":"3e3f37ad-2c6a-4abd-a980-19bfdf46a247","resolution":{"observed_at":"2026-08-03T19:24:39.349405Z","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":"2026-08-03T19:24:39.352263Z","title":"Transweather: Transformer-based restoration of im- ages degraded by adverse weather conditions","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":61,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.352263Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:87be6fb8af638a38b9e567f7627c26198357e38eb9b77556af3289ecbf8d9bf1","observation_id":"57dfbe95-4bdb-4dd8-8fbe-4882408549f7","resolution":{"observed_at":"2026-08-03T19:24:39.352263Z","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":"2026-08-03T19:24:39.355031Z","title":"Apisr: Anime production inspired real-world anime super-resolution","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":62,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.355031Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:ced018e8a8ad946bfc5984981e5df89938ca50bb2aeb1ed20f0b12576020d77d","observation_id":"4b79ff92-5b99-4c9c-ac49-54c7d345094e","resolution":{"observed_at":"2026-08-03T19:24:39.355031Z","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":"2026-08-03T19:24:39.357610Z","title":"Esrgan: En- hanced super-resolution generative adversarial networks","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":63,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.357610Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:ad634217e9fa7a9831b82f697fcd13196ea7733eb8a47ae8752bb5fc4b43928c","observation_id":"d281c04e-1475-4c0b-8ca7-70ac309e07ae","resolution":{"observed_at":"2026-08-03T19:24:39.357610Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1808.04560","last_updated":"2018-08-14T07:20:55Z","snapshot_observed_at":"2026-08-02T03:51:27.789282Z","submitted_at":"2018-08-14T07:20:55Z","title":"Deep Retinex Decomposition for Low-Light Enhancement","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1808.04560","snapshot_observed_at":"2026-08-03T19:24:39.360170Z","title":"Deep retinex decomposition for low-light enhancement","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":64,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.360170Z"},"links":{"cited_paper":"/paper/1808.04560","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:71b3b89e94b17891d3b816939f5ea5d0b7a4d9886b46b41aedf6b38b89814ce4","observation_id":"1e050e86-86c3-4e09-ba22-b49d222beb4e","resolution":{"observed_at":"2026-08-03T19:24:39.360170Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2505.24207","last_updated":"2025-05-30T04:36:52Z","snapshot_observed_at":"2026-08-07T12:27:33.267267Z","submitted_at":"2025-05-30T04:36:52Z","title":"Boosting All-in-One Image Restoration via Self-Improved Privilege Learning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2505.24207","snapshot_observed_at":"2026-08-03T19:24:39.363019Z","title":"Boosting all-in-one image restoration via self-improved priv- ilege learning.arXiv preprint arXiv:2505.24207, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":65,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.363019Z"},"links":{"cited_paper":"/paper/2505.24207","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:eda11ee726934b89022ed03295661398515644001cca0f51c997ec68a1625f2d","observation_id":"31796191-5a9b-46d5-8afe-d7a698f1d6c8","resolution":{"observed_at":"2026-08-03T19:24:39.363019Z","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":"2026-08-03T19:24:39.365893Z","title":"Cross-domain car detection model with integrated convolu- tional block attention mechanism.Image and Vision Com- puting, 140:104834, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":66,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.365893Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:a79e94721c2f928e8162bb867510441d2e1a4b5e8413545119c2a8fc8e0827b5","observation_id":"720a601b-3427-4f42-af86-2c69295ee113","resolution":{"observed_at":"2026-08-03T19:24:39.365893Z","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":"2026-08-03T19:24:39.368878Z","title":"Joint rain detection and removal from a single image with contextualized deep net- works.TPAMI, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":67,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.368878Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:9124505e190e629c916b75b99a6d1565214500b11fb3a4220a0da64e4adc421b","observation_id":"10fcfdd8-f0c0-45e5-8cb6-e37bd0542225","resolution":{"observed_at":"2026-08-03T19:24:39.368878Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2503.10615","last_updated":"2025-03-18T08:52:34Z","snapshot_observed_at":"2026-08-07T12:50:27.666060Z","submitted_at":"2025-03-13T17:56:05Z","title":"R1-Onevision: Advancing Generalized Multimodal Reasoning through Cross-Modal Formalization","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2503.10615","snapshot_observed_at":"2026-08-03T19:24:39.371464Z","title":"R1-onevision: Advancing gen- eralized multimodal reasoning through cross-modal formal- ization.arXiv preprint arXiv:2503.10615, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":68,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.371464Z"},"links":{"cited_paper":"/paper/2503.10615","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:8db8105da3d20ea176dd8ae83a14ab2b2d59d88bcf2616558c88984f52ef25ad","observation_id":"698a51c2-5987-4520-90f8-66e0c639e462","resolution":{"observed_at":"2026-08-03T19:24:39.371464Z","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":"2026-08-03T19:24:39.374648Z","title":"Uncertainty guided multi-scale residual learning-using a cycle spinning cnn for single image de-raining","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":69,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.374648Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:6fe478497ce0e6dd7fe3ddd67a6591d79501e893e8e089d90f5be095eb3219e9","observation_id":"b49ba789-08b4-4342-a4ea-ed38daeffc4a","resolution":{"observed_at":"2026-08-03T19:24:39.374648Z","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":"2026-08-03T19:24:39.377324Z","title":"Vla-r1: Enhancing rea- soning in vision-language-action models.arXiv preprint arXiv:2510.01623, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":70,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.377324Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:85373c30fe1c748440b2d2280be59a11c4e5ebd8ca7ad95fc017d86824808f24","observation_id":"3eaf8ee9-696a-41b0-a272-a03eaa87085b","resolution":{"observed_at":"2026-08-03T19:24:39.377324Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2511.17979","last_updated":"2026-06-30T16:15:01Z","snapshot_observed_at":"2026-08-03T20:54:24.309153Z","submitted_at":"2025-11-22T08:46:18Z","title":"FeRA: Frequency-Energy Constrained Routing for Effective Diffusion Adaptation Fine-Tuning","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2511.17979","snapshot_observed_at":"2026-08-03T19:24:39.380005Z","title":"Fera: Frequency-energy con- strained routing for effective diffusion adaptation fine- tuning.arXiv preprint arXiv:2511.17979, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":71,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.380005Z"},"links":{"cited_paper":"/paper/2511.17979","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:8b63c92ac914030d63415511ff7c988f9c3d50bd6d6e90cfec596edadc8bdae7","observation_id":"ebc83e44-7919-4f3f-b144-2ef19a304149","resolution":{"observed_at":"2026-08-03T19:24:39.380005Z","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":"2026-08-03T19:24:39.382892Z","title":"Multi-stage progressive image restoration","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":72,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.382892Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:546bdc2310ee5eb777eaa3e03d38c4c081b9077fa5a992625f0d6b2b5b74554e","observation_id":"c4e3fc4f-c29c-4ae9-ab5d-152f430efd88","resolution":{"observed_at":"2026-08-03T19:24:39.382892Z","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":"2026-08-03T19:24:39.385656Z","title":"Restormer: Efficient transformer for high-resolution image restoration","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":73,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.385656Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:2579ca742b6a1ff251c3ecd3b5e380a2c28286b21d23a85583ceb84877cccd8b","observation_id":"39d1a251-87a4-4915-82d7-016919c1676d","resolution":{"observed_at":"2026-08-03T19:24:39.385656Z","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":"2026-08-03T19:24:39.388380Z","title":"Learning enriched features for fast image restoration and enhancement.TPAMI, 2022","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":74,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.388380Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:b480351da3ccb479d9b9904f760ff15206d16939df70fe0d13c4a727c92f7c07","observation_id":"f70f38bf-3022-469b-976e-38e539f516ff","resolution":{"observed_at":"2026-08-03T19:24:39.388380Z","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":"2026-08-03T19:24:39.391306Z","title":"Density-aware single image de-raining using a multi-stream dense network","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":75,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.391306Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:a7038e06d8c08889ff2c8f90f9e26e0c906c6b69b1f5ac10add6de1466c881e9","observation_id":"ebb0888e-a5df-4799-896e-877de2bdf314","resolution":{"observed_at":"2026-08-03T19:24:39.391306Z","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":"2026-08-03T19:24:39.394231Z","title":"Ingredient-oriented multi- degradation learning for image restoration","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":76,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.394231Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:7093ef18c997cccf83fcb93cb34975aaa4f66e17f8aa5e502780cce817b3866d","observation_id":"cd26257a-c190-4665-846e-a822c7c7573f","resolution":{"observed_at":"2026-08-03T19:24:39.394231Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2503.12937","last_updated":"2025-08-04T04:22:09Z","snapshot_observed_at":"2026-08-06T21:34:54.534751Z","submitted_at":"2025-03-17T08:51:44Z","title":"R1-VL: Learning to Reason with Multimodal Large Language Models via Step-wise Group Relative Policy Optimization","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2503.12937","snapshot_observed_at":"2026-08-03T19:24:39.396917Z","title":"R1-vl: Learn- ing to reason with multimodal large language models via step-wise group relative policy optimization.arXiv preprint arXiv:2503.12937, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":77,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.396917Z"},"links":{"cited_paper":"/paper/2503.12937","citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:7a437684c9b6a13166cb7e116d0a2773855a6f10b63cdf347353268c8afa69e1","observation_id":"5cdd2b03-e76b-418c-aa8c-b937ce460b06","resolution":{"observed_at":"2026-08-03T19:24:39.396917Z","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":"2026-08-03T19:24:39.399893Z","title":"Beyond a gaussian denoiser: Residual learning of deep cnn for image denoising.TIP, 2017","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":78,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.399893Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:f0274e8d03ff49f783d0ed4975122743c73b7ea8da048d8c4cd97dd1cf488673","observation_id":"ba4e89bd-6b3a-4286-ba60-e370fa7c969b","resolution":{"observed_at":"2026-08-03T19:24:39.399893Z","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":"2026-08-03T19:24:39.402677Z","title":"Learning deep CNN denoiser prior for image restoration","venue":null,"work_id":null,"year":2017},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":79,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.402677Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:e4da08744dbc2565e2dc5ba9c700ddee9d4d077c2fdb4771e2e11af126d16fa0","observation_id":"55ed92a6-aa8a-4ce4-86d4-fd8e6ebbbaa7","resolution":{"observed_at":"2026-08-03T19:24:39.402677Z","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":"2026-08-03T19:24:39.405419Z","title":"Ffdnet: Toward a fast and flexible solution for cnn-based image denoising","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":80,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.405419Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:55aed95f30b94b24038fc9a129e16d77cf1d67f4f53341d95343a3c3963d5fb1","observation_id":"70c5ca71-622a-45c4-95f5-579ff748bad4","resolution":{"observed_at":"2026-08-03T19:24:39.405419Z","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":"2026-08-03T19:24:39.408449Z","title":"The unreasonable effectiveness of deep features as a perceptual metric","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":81,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.408449Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:7fb82247d1dd5bbf0a9334c906f0b7df8671de0d35df7dc195eabe06611a2419","observation_id":"9f3cd35b-f4e5-4146-b3d4-03efebe9ad37","resolution":{"observed_at":"2026-08-03T19:24:39.408449Z","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":"2026-08-03T19:24:39.411387Z","title":null,"venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":82,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.411387Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:0a71749d566817d8d6b13701754b5a298bae6cf9ea170433708d3df4d59ddb3e","observation_id":"3e75ad3c-7547-46d3-a9ae-fc620b2cdff9","resolution":{"observed_at":"2026-08-03T19:24:39.411387Z","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":"2026-08-03T19:24:39.414977Z","title":"Real-world remote sensing image dehaz- ing: Benchmark and baseline.IEEE Transactions on Geo- science and Remote Sensing, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":83,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.414977Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:21d718e7a2f04c3a5e33d0c88866cf049eaec6d6d76e8040edc4407e8f44b2a9","observation_id":"ed38011c-d8a3-4483-8138-92e7123611f3","resolution":{"observed_at":"2026-08-03T19:24:39.414977Z","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":"2026-08-03T19:24:39.417866Z","title":"A robustly optimized bert pre-training approach with post-training","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":84,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.417866Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:b609e156d66ea2ca4ce93252aee6de2b7df612581c2651c5889d6af6544dc8d8","observation_id":"b07e3c5a-b33f-4f0e-8711-80d04b2f9cf6","resolution":{"observed_at":"2026-08-03T19:24:39.417866Z","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":"2026-08-03T19:24:39.420904Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":85,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.420904Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:dda849eb15136c3a93b6b434b79e0bbb53ed564903d54aafef11286c89007715","observation_id":"3c460d4c-b3ac-4c40-89cd-d3b3bb907776","resolution":{"observed_at":"2026-08-03T19:24:39.420904Z","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":"2026-08-03T19:24:39.423536Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":86,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.423536Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:b2d600b4c591518b960cde6aa9762680ff28bdf1fe46e9538845147af33c2466","observation_id":"2480ac6b-b8a6-4a91-b599-500b4a0f7293","resolution":{"observed_at":"2026-08-03T19:24:39.423536Z","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":"2026-08-03T19:24:39.426299Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":87,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.426299Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:d5bb177658f4e5ed895f3683f38e1c88b93e02822c77ac44bd6056cab5749f2b","observation_id":"18e7216c-2258-4f9b-9a11-9d04196c0bfe","resolution":{"observed_at":"2026-08-03T19:24:39.426299Z","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":"2026-08-03T19:24:39.428968Z","title":"Consider categories such as denoising (0/1/2, different noise levels), deraining (3), dehaz- ing (4), deblurring (5), or low-light enhancement (6)","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":88,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.428968Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:a724657887f4c491593c89e992f7568ddd00a21c0c34b7d7450f8e09e01b0630","observation_id":"4c0c97c6-c61e-4085-861d-4d2fb99c099d","resolution":{"observed_at":"2026-08-03T19:24:39.428968Z","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":"2026-08-03T19:24:39.431602Z","title":"Pay attention to: • Noise or streak removal quality for denois- ing/deraining","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":89,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.431602Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:624e626f5a956bd91f173a007a24ade1a2ed4f9c03df079d68c3abcca577af6d","observation_id":"10bb0c0f-0014-4b90-a713-88c0c4e93167","resolution":{"observed_at":"2026-08-03T19:24:39.431602Z","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":"2026-08-03T19:24:39.434201Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":90,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.434201Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:a4eefcf5ceea135ea18a4c9c00ca78e1ee269119efc9d21d03ae75a39aeec120","observation_id":"997ba2b7-131f-41d7-b6c7-08bf80a295ac","resolution":{"observed_at":"2026-08-03T19:24:39.434201Z","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":"2026-08-03T19:24:39.437108Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO","version":3},"reference_index":91,"source":"pdf_text","source_observed_at":"2026-08-03T19:24:39.437108Z"},"links":{"citing_paper":"/paper/2512.00814"},"observation_digest":"sha256:4f9aa1fab47bfe1aeeb4b4c225d7aa52c04a252f859fd0de7f061871fdddf30c","observation_id":"b23c2086-8dcd-42ea-a965-738ecee70657","resolution":{"observed_at":"2026-08-03T19:24:39.437108Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2512.00814","last_updated":"2026-05-27T04:00:29Z","latest_version":3,"primary_category":"cs.CV","snapshot_observed_at":"2026-08-06T08:11:35.009189Z","submitted_at":"2025-11-30T09:42:24Z","title":"IRPO: Boosting Image Restoration via Post-training GRPO"},"reference_resolution":{"displayed":91,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":91,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":91},"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 9 August 2026, this Paper Citation Record lists 91 of 91 outbound references and 1 inbound Pith citation observation for arXiv:2512.00814."}