{"as_of":"2026-08-19T05:09:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:cc62a1262c30a7c9fc6b04424f04d37784a75fcbcae3f90fe337ae2ffdbd2629","coverage":[{"denominator":56,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":56,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-01T19:16:45.202633Z","state":"measured"},{"denominator":57,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":57,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-18T06:34:40.430872+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":"2026-08-06T00:08:25.571429Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"pith","source_observed_at":"2026-08-06T00:08:28.562083Z","state":"measured"}],"external_citation_measurements":[],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"cited_work":{"arxiv_id":"2607.17043","doi":null,"metadata_source":"pith","pith_arxiv_id":"2607.17043","snapshot_observed_at":"2026-08-06T00:08:28.562083Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","venue":"cs.CL","work_id":"19354def-1367-452a-940b-c2f3dc67673f","year":2026},"citing_paper":{"arxiv_id":"2608.01522","last_updated":"2026-08-02T22:19:01Z","snapshot_observed_at":"2026-08-15T16:34:06.458991Z","submitted_at":"2026-08-02T22:19:01Z","title":"Question Begets Question: Self-Evolving Curriculum for Reinforcement Fine-Tuning on Competition Mathematics","version":1},"reference_index":28,"source":"arxiv_source","source_observed_at":"2026-08-06T00:08:25.571429Z"},"links":{"cited_paper":"/paper/2607.17043","citing_paper":"/paper/2608.01522"},"observation_digest":"sha256:d52db635fb18b2c335badd5183c2fbce357179ef27f34a9c38d1e1d8379aa27c","observation_id":"fdb61e68-0c05-42de-8a15-7b65e0b3a022","resolution":{"observed_at":"2026-08-06T00:08:28.703155Z","resolver_source":"local_arxiv","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-18T06:34:40.430872+00:00","source":"crossref"},{"observed_at":"2026-08-18T06:34:34.496301+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2607.17043/citation-record","integrity":"/paper/2607.17043/integrity","json":"/paper/2607.17043/citation-record.json","paper":"/paper/2607.17043"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T19:16:37.800084Z","title":"2024 , eprint=","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":1,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:37.800084Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:831e6eb279097dcadf1ba29b1c58507bfbc555313908d80cec988273f567a401","observation_id":"7ba904ad-e4b8-4849-8c3f-7bd4a74345fa","resolution":{"observed_at":"2026-08-01T19:16:37.800084Z","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":"2026-08-01T19:16:37.911031Z","title":"Forty-first International Conference on Machine Learning , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":2,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:37.911031Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:ba7b7db7b89671b8f6517b31997fd8fe60e6d3aa26416d06608e76f0661798fc","observation_id":"bca8b6fe-7418-4163-b894-749d29777dd4","resolution":{"observed_at":"2026-08-01T19:16:37.911031Z","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":"2026-08-01T19:16:37.972344Z","title":"First Conference on Language Modeling , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":3,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:37.972344Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:12b3ecb8668bb6926ac0915d8eb86f84e593140c557a10938ff4e43ea022779e","observation_id":"8a38c97e-fc3f-4c17-92f2-ed71ff1dbb50","resolution":{"observed_at":"2026-08-01T19:16:37.972344Z","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":"2026-08-01T19:16:38.050986Z","title":"Forty-second International Conference on Machine Learning , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":4,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.050986Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:178f5c2994b1ea3ac345ec67dec7173d7509d78bfa8261bc819715acd08b9ae0","observation_id":"fae68403-0f18-45f8-8149-12ea99187812","resolution":{"observed_at":"2026-08-01T19:16:38.050986Z","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":"2026-08-01T19:16:38.149880Z","title":"The Thirteenth International Conference on Learning Representations , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":5,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.149880Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:91615669beb9c36512df372b59ba33a9777e76dd3152b7ad55d455b6255a2669","observation_id":"cd7f2cbb-a295-46c6-82d6-7cba6dcb6ed7","resolution":{"observed_at":"2026-08-01T19:16:38.149880Z","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":"2026-08-01T19:16:38.159710Z","title":"Machine-generated text detection prevents language model collapse","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":6,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.159710Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:311014788b85f14dbd2f18269defa4c62ae138edea854b26f7b67f14c460551a","observation_id":"2f5966b1-a0a3-407f-bf34-1a2fe3205fdf","resolution":{"observed_at":"2026-08-01T19:16:38.159710Z","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":"2026-08-01T19:16:38.163184Z","title":"The Thirty-ninth Annual Conference on Neural Information Processing Systems , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":7,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.163184Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:52fd9e0357b0aa9694d5a17e5c9175606a24f1ac09fd3e2dd20f09953a662f71","observation_id":"c6233ec5-a28c-473f-8d07-e2c6762789df","resolution":{"observed_at":"2026-08-01T19:16:38.163184Z","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":"2026-08-01T19:16:38.204167Z","title":"First Conference on Language Modeling , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":8,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.204167Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:80bf2b934da9a72ee66f0aff301a38a2373b4780eb376047703f59dc37955b4b","observation_id":"626a08a3-d8b1-4890-8334-dd84845bbc8d","resolution":{"observed_at":"2026-08-01T19:16:38.204167Z","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":"2026-08-01T19:16:38.355848Z","title":"Generating Datasets with Pretrained Language Models","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":9,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.355848Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:4a9030425704ddedd093de34e7d0e5dcb70b4b49b3482ee281e6289c363185dc","observation_id":"99094271-036a-4d80-8e07-8773c93c8608","resolution":{"observed_at":"2026-08-01T19:16:38.355848Z","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":"10.18653/v1/2021.emnlp-main.562","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Synthetic Data Augmentation for Zero-Shot Cross-Lingual Question Answering","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","work_id":"e47fbfef-d9cc-4e8d-95ff-4f21dd019511","year":2021},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":10,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.492939Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:8b9b3534b48e3eeb22c75a7c564be9e32d64b908c5a9dcc411fee7690f65e5c6","observation_id":"f28b616d-6477-4835-820d-dfec94b9eeb0","resolution":{"observed_at":"2026-08-01T19:19:08.461907Z","resolver_source":"doi","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-18T06:34:40.430872+00:00","source":"crossref"},{"observed_at":"2026-08-18T06:34:34.496301+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T19:16:38.666561Z","title":"WildChat: 1M Chat","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":11,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.666561Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:87c9200f6ef643684cef8f0e1c6806eba25868f1719d8b40ea0fd133cb95669e","observation_id":"f0bac9a9-b2de-43f7-a1a7-caf472d095ec","resolution":{"observed_at":"2026-08-01T19:16:38.666561Z","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":"2026-08-01T19:16:38.836667Z","title":"Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":12,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:38.836667Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:da5c9901c53be0e6ab9f6ae5cab188720ace1bb5043764c40c93ffc935251660","observation_id":"fa8bb71e-d67c-4b24-9f7f-1e07f7239b00","resolution":{"observed_at":"2026-08-01T19:16:38.836667Z","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":"2026-08-01T19:16:39.005431Z","title":"2025 , eprint=","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":13,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:39.005431Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:02768e25e8a628687d6f6799eff27c853f586315793f67a50e35b67a87462fc5","observation_id":"ca4285d4-ef78-4987-a84f-aee48a6b3c44","resolution":{"observed_at":"2026-08-01T19:16:39.005431Z","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":"2026-08-01T19:16:39.206081Z","title":"2025 , eprint=","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":14,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:39.206081Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:3e2c1c304087292640f96d3357ea7c4c08fd0159a4bb660337f2e591b18ed54b","observation_id":"a217a4a5-b4a8-436a-8400-6bbb0365ecb5","resolution":{"observed_at":"2026-08-01T19:16:39.206081Z","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":"2026-08-01T19:16:39.424723Z","title":"2024 , eprint=","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":15,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:39.424723Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:6b9e069842af7f6b0e9027867ee662c9c1aa1821e00be3bfcc6eb0ef3a580d41","observation_id":"ca64a94e-e342-4ac6-bf38-32cfeb644658","resolution":{"observed_at":"2026-08-01T19:16:39.424723Z","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":"2026-08-01T19:16:39.593123Z","title":"ChemOrch: Empowering","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":16,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:39.593123Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:875d8b875a9a28776b466dd595ed31f7bec45f924bb910c0e469d3ba17e556a3","observation_id":"b3ac5142-2f97-4ffb-848c-35e99c597ef0","resolution":{"observed_at":"2026-08-01T19:16:39.593123Z","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":"2026-08-01T19:16:39.858154Z","title":"AugGPT: Leveraging ChatGPT for Text Data Augmentation , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":17,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:39.858154Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:5436db5b1c9988d98575db35ee56e31aa89c06d64d992ef643cd24609d7f038f","observation_id":"a997aa47-cd08-4956-a287-f810908393f2","resolution":{"observed_at":"2026-08-01T19:16:39.858154Z","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":"2026-08-01T19:16:40.069628Z","title":"MetaSynth: Meta-Prompting-Driven Agentic Scaffolds for Diverse Synthetic Data Generation , url=","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":18,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:40.069628Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:d9f70d70681885c9a0e10e45a88dd9fa1c852208be9b258f8418584989fe57c7","observation_id":"6670701a-8289-4d16-8c1d-8f111c21251b","resolution":{"observed_at":"2026-08-01T19:16:40.069628Z","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":"2026-08-01T19:16:40.231652Z","title":"The Thirty-eighth Annual Conference on Neural Information Processing Systems , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":19,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:40.231652Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:122389738c4b8c0257f2f08bcdec23d075419df5a7938286b1435664f0904458","observation_id":"7056f156-31f4-439b-be41-cdede75f3a93","resolution":{"observed_at":"2026-08-01T19:16:40.231652Z","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":"2026-08-01T19:16:40.407982Z","title":"The Thirteenth International Conference on Learning Representations , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":20,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:40.407982Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:1eda856bd45eb081d440701dcf104def34dc021f48b523436bc1396a0be00529","observation_id":"d7416383-9d9b-40b3-980b-72286b84c52a","resolution":{"observed_at":"2026-08-01T19:16:40.407982Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2110.14168","last_updated":"2021-11-18T00:23:45Z","snapshot_observed_at":"2026-08-14T02:43:01.480086Z","submitted_at":"2021-10-27T04:49:45Z","title":"Training Verifiers to Solve Math Word Problems","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2110.14168","snapshot_observed_at":"2026-08-01T19:16:40.583129Z","title":"arXiv preprint arXiv:2110.14168 , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":21,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:40.583129Z"},"links":{"cited_paper":"/paper/2110.14168","citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:bb8ee00d7bf9d8f004aa5389ea2f5859b4c346920fe7760fd8ba7aba21e7199a","observation_id":"c95b31ad-066e-48cd-9a93-8ef3e4c7d046","resolution":{"observed_at":"2026-08-01T19:16:40.583129Z","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":"2026-08-01T19:16:40.811463Z","title":"Journal of Educational and Behavioral Statistics , volume =","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":22,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:40.811463Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:fbaba5d0de667d7b33413b0385886ed313e2c4afd6e96393e2c7a6be2c1319b9","observation_id":"8631f0b4-67bb-41b9-b9c8-8e59b581e471","resolution":{"observed_at":"2026-08-01T19:16:40.811463Z","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":"2026-08-01T19:16:40.957900Z","title":"2024 , url=","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":23,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:40.957900Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:e78c90e464f144d99d88197e5a403504b73f37dceef57dbbc2f0388aa3f7401c","observation_id":"4028f733-04c8-42a6-8bb2-8c6cca0480c0","resolution":{"observed_at":"2026-08-01T19:16:40.957900Z","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":"2026-08-01T19:16:41.123251Z","title":"2021 , eprint=","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":24,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:41.123251Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:fba63c47bef11df2d79b2b49cb9443780ec6b07823d7924da5b9e6c4b9c69c21","observation_id":"93907abb-ec6c-4519-9154-dca22124392e","resolution":{"observed_at":"2026-08-01T19:16:41.123251Z","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":"2026-08-01T19:16:41.291523Z","title":"2021 , eprint=","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":25,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:41.291523Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:349b296fb329f7b16d923007715549054c7a2b626d64fdface174fbfcc2e5ffb","observation_id":"4e24c473-e491-48da-a4f8-a39898762a44","resolution":{"observed_at":"2026-08-01T19:16:41.291523Z","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":"2026-08-01T19:16:41.408086Z","title":"Bowman , booktitle=","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":26,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:41.408086Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:6b16ca60e43d62da74b440f1719296ca1962c89591d59e9a934581ae0d5bca30","observation_id":"e1ed80a8-a2ec-4f8a-ac0e-ffb73a4d88aa","resolution":{"observed_at":"2026-08-01T19:16:41.408086Z","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":"2026-08-01T19:16:41.500661Z","title":"2023 , eprint=","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":27,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:41.500661Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:44a40c74c8fe63f04be25e028ecbe10acd57ff658083a32fb2082d34120c7446","observation_id":"30446ef8-0586-46d2-9368-1b2237f3022f","resolution":{"observed_at":"2026-08-01T19:16:41.500661Z","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":"2026-08-01T19:16:41.628791Z","title":"Kernel Language Entropy: Fine-grained Uncertainty Quantification for","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":28,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:41.628791Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:3f52357af20fe7227939cb4503672e21fd27578ebbdd3c282c4aeb4797093649","observation_id":"f04fb078-a386-4b4b-a5c9-e810ef619d08","resolution":{"observed_at":"2026-08-01T19:16:41.628791Z","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":"2026-08-01T19:16:41.748959Z","title":null,"venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":29,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:41.748959Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:ae11c4bb52109948e16c8ea1940c0684522a8a4f31fa28bc5f9022d02723da2b","observation_id":"e6d7c491-2be7-4089-8431-8beeb49b1b43","resolution":{"observed_at":"2026-08-01T19:16:41.748959Z","resolver_source":null,"status":"parse_uncertain"},"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":"2026-08-01T19:16:41.909186Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":30,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:41.909186Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:0c88569b2d315f90c7faebd08bfa4b4d4b4046e46652b6123e82359c0ff9875b","observation_id":"3e284aa7-81da-4a4e-aa8a-40df3e75cde0","resolution":{"observed_at":"2026-08-01T19:16:41.909186Z","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":"2026-08-01T19:16:42.069626Z","title":"2025 , eprint=","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":31,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:42.069626Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:7311f5cb1ce17a15728019c0e044558941c10560734c0c4f8104cf568ac50da8","observation_id":"c8d58358-543d-4532-82e0-c300eca4537c","resolution":{"observed_at":"2026-08-01T19:16:42.069626Z","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":"2026-08-01T19:16:42.241620Z","title":"2024 , url =","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":32,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:42.241620Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:a73c2b23523859a2d65372db8d91abbf92ca0618e6c57f6f118f6a9306de791a","observation_id":"45d0a908-5848-4497-b1a0-4a57dd1ad673","resolution":{"observed_at":"2026-08-01T19:16:42.241620Z","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":"2026-08-01T19:16:42.358076Z","title":"Gemma 3 , url=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":33,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:42.358076Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:42259884e0b87befa4444d992efe2b165fa729e86b7f5e25f0911cabea6d9a21","observation_id":"3d3d9696-fd1d-4f5f-9888-2c9749005dd4","resolution":{"observed_at":"2026-08-01T19:16:42.358076Z","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":"2026-08-01T19:16:42.474509Z","title":"Forty-second International Conference on Machine Learning , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":34,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:42.474509Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:15e245f2f15a4df92548f6f7ef449cbbd68a9f4456f9649ebbe92309bb711a12","observation_id":"e3d53f09-c6ad-4b3d-963f-df3b0e562185","resolution":{"observed_at":"2026-08-01T19:16:42.474509Z","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":"2026-08-01T19:16:42.646017Z","title":"The Thirteenth International Conference on Learning Representations , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":35,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:42.646017Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:9bbfb73ea8b7035514fc65a449334a19dd0b015ca50f61a560818f0cbb3777fc","observation_id":"29367883-9448-4fbe-85e2-c27807c7cde3","resolution":{"observed_at":"2026-08-01T19:16:42.646017Z","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":"2026-08-01T19:16:42.897469Z","title":"2023 , eprint=","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":36,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:42.897469Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:0ecf1dbb6b9c8cdeee8699bfcc71ac0b50e230175c4359eeaa0fe0e2027d78dd","observation_id":"d21b5bec-a08b-4ec4-99ee-e16b7882f558","resolution":{"observed_at":"2026-08-01T19:16:42.897469Z","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":"2026-08-01T19:16:42.934129Z","title":"and Le, Quoc V and Firat, Orhan","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":37,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:42.934129Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:ed3ea0d0c9dcd909fd6a5be22e04f07d08c1ad60c626811db3c6e6fee7aa4c67","observation_id":"315b6c2f-43c8-46b1-b6aa-bbbf882473d4","resolution":{"observed_at":"2026-08-01T19:16:42.934129Z","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":"2026-08-01T19:16:42.956431Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":38,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:42.956431Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:2a0b508b5f1f43af0cce177600ca5f28ba441a110ecd9c5f2b20764f89af9a5d","observation_id":"e19134b3-4931-4d58-872e-0e2db43445c5","resolution":{"observed_at":"2026-08-01T19:16:42.956431Z","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":"2026-08-01T19:16:43.101368Z","title":"2025 , eprint=","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":39,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:43.101368Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:8cfd12773172d2cdb0cd2de8ab136e743f7bf5899c6d8f14e6cedf3553e0bf0d","observation_id":"43913c31-1487-4054-9d3c-fad1e071fa57","resolution":{"observed_at":"2026-08-01T19:16:43.101368Z","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":"2026-08-01T19:16:43.211634Z","title":"L lama F actory: Unified Efficient Fine-Tuning of 100+ Language Models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":40,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:43.211634Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:179a33c2f11c30f8f8b0c057d5d0249e6e3883729d84c9543c41b7ddb4af87f6","observation_id":"fc40a455-6f63-40f5-950e-6f1f3cf030a9","resolution":{"observed_at":"2026-08-01T19:16:43.211634Z","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":"2026-08-01T19:16:43.354458Z","title":"2016 , eprint=","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":41,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:43.354458Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:8510b84cc179ddf4225ac3db13f93ff13e42c410137e45c9a0f68a6173b555a4","observation_id":"fcac6cd5-998f-41e1-a8ac-62e56686453b","resolution":{"observed_at":"2026-08-01T19:16:43.354458Z","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":"2026-08-01T19:16:43.608554Z","title":"The Thirty-eighth Annual Conference on Neural Information Processing Systems , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":42,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:43.608554Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:ae60cd7ed1f90627770cacd0dc8b6ab636d28b592a6c124c24281fdbe1225508","observation_id":"9fd482c3-8e75-4c60-a3ab-8fe8a7d6a207","resolution":{"observed_at":"2026-08-01T19:16:43.608554Z","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":"10.18653/v1/2025.findings-acl.439","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"CDS : Data Synthesis Method Guided by Cognitive Diagnosis Theory","venue":null,"work_id":"906d1251-65ac-4908-b887-ac353fbc44cf","year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":43,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:43.776350Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:51d42e0c0bdc0fd6e9b3db91331f5013f7e3bca73d01d4a9018c55428569cb41","observation_id":"498673f2-4e28-449d-a2fd-f3ba3e9aa386","resolution":{"observed_at":"2026-08-01T19:19:08.279888Z","resolver_source":"doi","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-18T06:34:40.430872+00:00","source":"crossref"},{"observed_at":"2026-08-18T06:34:34.496301+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T19:16:43.951194Z","title":"2024 , eprint=","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":44,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:43.951194Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:258fa7bc4bf9fa810f4d89e28c7925548608ca8b5543f96410dda89c1093a486","observation_id":"a252d8d6-25ec-43a2-9293-c9d5f5eaf16c","resolution":{"observed_at":"2026-08-01T19:16:43.951194Z","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":"2026-08-01T19:16:44.125647Z","title":"2022 , eprint=","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":45,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:44.125647Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:fa89369abaac1a1e09d1a765be2b2d4f0b456464d253b4e41ef29436746b46d2","observation_id":"dcd0e12c-fc1f-4bae-8045-50d9a3fbf6ca","resolution":{"observed_at":"2026-08-01T19:16:44.125647Z","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":"2026-08-01T19:16:44.300798Z","title":"2025 , eprint=","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":46,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:44.300798Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:92d75f0137233158450266479f8a938680473ff063eebd0a9fea8e8421257c5a","observation_id":"926a7241-20c8-4328-94d9-048cd20c85d9","resolution":{"observed_at":"2026-08-01T19:16:44.300798Z","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":"2026-08-01T19:16:44.503636Z","title":"2023 , eprint=","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":47,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:44.503636Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:3e827a1513218687a96237467db11e53343ba29fddbb22ffe046959e628e96d0","observation_id":"f01c923f-3201-4a7e-a2fd-97556fee14fe","resolution":{"observed_at":"2026-08-01T19:16:44.503636Z","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":"2026-08-01T19:16:44.623575Z","title":"2026 , eprint=","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":48,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:44.623575Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:a4013bf57526e1f5f0733497f66efff0f6b92f8de9af7c4941ccb4424e54511c","observation_id":"95634d21-f569-4bb9-a16b-e79839860ab8","resolution":{"observed_at":"2026-08-01T19:16:44.623575Z","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":"2026-08-01T19:16:44.693087Z","title":"2026 , eprint=","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":49,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:44.693087Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:f8d8ce6f611329e9b3ffdd733fc6687370665100954657dce943ba3e00c2f1c0","observation_id":"9be0d15b-38dc-4cfb-948f-282b0db9836a","resolution":{"observed_at":"2026-08-01T19:16:44.693087Z","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":"10.18653/v1/2025.naacl-long.317","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Language Models can Categorize System Inputs for Performance Analysis","venue":null,"work_id":"ca7d49b8-48de-4b75-8441-308731722a17","year":2025},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":50,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:44.775223Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:c951d09ce8e6bab25bfe50d849f5b65af299f487a01164fb383329b13d81e3de","observation_id":"23299422-824f-45a5-9c98-cce71f437d31","resolution":{"observed_at":"2026-08-01T19:19:08.176100Z","resolver_source":"doi","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-18T06:34:40.430872+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-18T06:34:40.430872+00:00","source":"crossref"},{"observed_at":"2026-08-18T06:34:34.496301+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-08-01T19:16:44.858916Z","title":"Second Conference on Language Modeling , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":51,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:44.858916Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:183537137239bdf0cb70d60eee03e87093fb81b37f100dfcaa614bc3132c7ff2","observation_id":"bf555187-8dad-468f-a47f-1d41298f8c52","resolution":{"observed_at":"2026-08-01T19:16:44.858916Z","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":"2026-08-01T19:16:44.941979Z","title":"Forty-second International Conference on Machine Learning , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":52,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:44.941979Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:450f27b67f51997d4059eda5b31b4861e2d8e60173f46c772685776ba84bf2a9","observation_id":"d540051b-933f-4a00-8edc-86995baa9d94","resolution":{"observed_at":"2026-08-01T19:16:44.941979Z","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":"2026-08-01T19:16:45.002164Z","title":"2024 , eprint=","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":53,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:45.002164Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:5c48a9e4cad08ecec2b2efdfab0275f27465c157a8e679890cc844a646d18238","observation_id":"e55852d6-264d-4ce1-8575-5e0dc82ce93f","resolution":{"observed_at":"2026-08-01T19:16:45.002164Z","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":"2026-08-01T19:16:45.022515Z","title":"Superfiltering: Weak-to-Strong Data Filtering for Fast Instruction-Tuning","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":54,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:45.022515Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:97d9fce2d5bd6315ded09a550c4a31c9e64a83915cf290bccaaeb795bb9c3082","observation_id":"5f9a9295-8fd1-4cc2-9c0c-88087406b018","resolution":{"observed_at":"2026-08-01T19:16:45.022515Z","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":"2026-08-01T19:16:45.044690Z","title":"The Fourteenth International Conference on Learning Representations , year=","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":55,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:45.044690Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:8396d26fd5296f46669af73866876065d8ce43e1e805eae3bdf44dcd74ee83f2","observation_id":"95a6c0ba-16a4-47b6-9057-920769197bbf","resolution":{"observed_at":"2026-08-01T19:16:45.044690Z","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":"2026-08-01T19:16:45.202633Z","title":"2026 , eprint=","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning","version":1},"reference_index":56,"source":"arxiv_source","source_observed_at":"2026-08-01T19:16:45.202633Z"},"links":{"citing_paper":"/paper/2607.17043"},"observation_digest":"sha256:732062eea2ed3c3e611ce62db05957b5000b4ddd8b7087dac84e091f23a99cc6","observation_id":"d2a905b5-e227-464e-a956-4651fb5eddbd","resolution":{"observed_at":"2026-08-01T19:16:45.202633Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2607.17043","last_updated":"2026-07-19T03:20:01Z","latest_version":1,"primary_category":"cs.CL","snapshot_observed_at":"2026-08-16T00:43:31.786037Z","submitted_at":"2026-07-19T03:20:01Z","title":"Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning"},"reference_resolution":{"displayed":56,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":1,"unresolved":52,"verified_exact":3,"verified_fuzzy":0},"total_outbound_references":56},"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":"2026-08-18T06:34:40.430872+00:00","source":"crossref"},{"observed_at":"2026-08-18T06:34:34.496301+00:00","source":"retraction_watch"}],"thesis":"As of 19 August 2026, this Paper Citation Record lists 56 of 56 outbound references and 1 inbound Pith citation observation for arXiv:2607.17043."}