{"as_of":"2026-08-10T00:26:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:247cdf1ae65401cd6bc2dd00525b2fd5b35c2c65156a46a387be8ab34cdaacfd","coverage":[{"denominator":0,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":14,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":14,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-09T06:31:02.800959+00:00","state":"measured"},{"denominator":14,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":14,"source":"paper_references, paper_reference_links","source_observed_at":"2026-08-08T18:54:32.353949Z","state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":1,"source":"arxiv_reference","source_observed_at":"2026-08-05T02:28:24.338817Z","state":"measured"}],"external_citation_measurements":[{"count":8,"observed_at":"2026-08-05T02:28:24.338817Z","source":"arxiv_reference"}],"inbound":[{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":"2305.09246","doi":"10.48550/arxiv.2305.09246","metadata_source":"arxiv_reference","pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2305.09246 , year=","venue":"arXiv (Cornell University)","work_id":"8815fb68-5998-4c8e-aa14-382ab7927214","year":2023},"citing_paper":{"arxiv_id":"2307.06435","last_updated":"2024-10-17T01:10:40Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2023-07-12T20:01:52Z","title":"A Comprehensive Overview of Large Language Models","version":10},"reference_index":184,"source":"pdf_text","source_observed_at":"2026-05-19T20:28:38.900026Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2307.06435"},"observation_digest":"sha256:a8b40cfa41ddadd6d93f76d2313ff46f88a64f3d001c9507431e3878bf76aaec","observation_id":"a17b9d42-079a-4054-9af9-99bd1de6bf93","resolution":{"observed_at":"2026-05-19T20:28:39.584384Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":"2305.09246","doi":"10.48550/arxiv.2305.09246","metadata_source":"arxiv_reference","pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2305.09246 , year=","venue":"arXiv (Cornell University)","work_id":"8815fb68-5998-4c8e-aa14-382ab7927214","year":2023},"citing_paper":{"arxiv_id":"2410.20791","last_updated":"2026-04-06T19:29:25Z","snapshot_observed_at":"2026-08-03T01:41:19.546733Z","submitted_at":"2024-10-28T07:16:00Z","title":"From Cool Demos to Production-Ready FMware: Core Challenges and a Technology Roadmap","version":3},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-05-23T19:07:21.016824Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2410.20791"},"observation_digest":"sha256:4beaa6e775d489980ee9f9668006cfd7802ea5726828f7b3b1972503976a88f4","observation_id":"3b3810f5-0a64-4834-a2fa-fc7cec50e540","resolution":{"observed_at":"2026-05-23T19:08:20.935697Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-08T18:54:32.353949Z","title":null,"venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2502.05562","last_updated":"2025-02-08T13:21:34Z","snapshot_observed_at":"2026-08-08T18:47:29.126411Z","submitted_at":"2025-02-08T13:21:34Z","title":"Can Large Language Models Be Query Optimizer for Relational Databases?","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-08-08T18:54:32.353949Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2502.05562"},"observation_digest":"sha256:c7111c823eb1b8cc53cabf4fa214d36dd60254b3b793045e09de20dcb6b188f1","observation_id":"8799ccc2-9085-41d3-9fbe-10ed33cfcfc9","resolution":{"observed_at":"2026-08-08T18:54:32.353949Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-07T14:48:02.545145Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.17632","last_updated":"2025-05-23T08:45:46Z","snapshot_observed_at":"2026-08-08T19:01:21.470738Z","submitted_at":"2025-05-23T08:45:46Z","title":"ReqBrain: Task-Specific Instruction Tuning of LLMs for AI-Assisted Requirements Generation","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-08-07T14:48:02.545145Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2505.17632"},"observation_digest":"sha256:b0f00a2d2d8549e867259bd48666cb13b2ca46aa40f006a29ded8edd6029c6d9","observation_id":"4df64c3e-463d-4036-8ef3-d208509a3a8f","resolution":{"observed_at":"2026-08-07T14:48:02.545145Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-07T14:31:28.279489Z","title":null,"venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2505.18799","last_updated":"2025-06-18T05:56:01Z","snapshot_observed_at":"2026-08-09T07:37:21.236571Z","submitted_at":"2025-05-24T17:19:34Z","title":"ALPS: Attention Localization and Pruning Strategy for Efficient Alignment of Large Language Models","version":4},"reference_index":7,"source":"arxiv_source","source_observed_at":"2026-08-07T14:31:28.279489Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2505.18799"},"observation_digest":"sha256:0a92a03490bd6aa970175e204005a62b2295e4d4faab1e3c2ad436e0ec083daf","observation_id":"7fcd1e42-e9fc-4010-b60f-aae0e4afd8d4","resolution":{"observed_at":"2026-08-07T14:31:28.279489Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-07T04:55:35.255414Z","title":"Maybe only 0.5% data is needed: A preliminary exploration of low training data instruction tuning","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.09329","last_updated":"2025-06-11T02:08:52Z","snapshot_observed_at":"2026-08-09T02:50:04.519855Z","submitted_at":"2025-06-11T02:08:52Z","title":"Towards Efficient and Effective Alignment of Large Language Models","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-08-07T04:55:35.255414Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2506.09329"},"observation_digest":"sha256:e68801a8445716b33ad668c81765553d78f7d5d0b687ac6e39fa0c02ec6074db","observation_id":"f5508cc0-8a67-49c5-9e72-5f3b9356a530","resolution":{"observed_at":"2026-08-07T04:55:35.255414Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-07T12:40:25.256358Z","title":"Maybe only 0.5% data is needed: A preliminary exploration of low training data instruction tuning.arXiv preprint arXiv:2305.09246, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2506.15702","last_updated":"2025-05-30T01:54:12Z","snapshot_observed_at":"2026-08-08T01:32:10.806688Z","submitted_at":"2025-05-30T01:54:12Z","title":"Minifinetuning: Low-Data Generation Domain Adaptation through Corrective Self-Distillation","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-08-07T12:40:25.256358Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2506.15702"},"observation_digest":"sha256:580869e11ee413bab27e056b6316c9a0bc7a81d909b356cce6fbd851c89f967b","observation_id":"a36c054a-8148-4258-b397-d8431c44bc66","resolution":{"observed_at":"2026-08-07T12:40:25.256358Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-04T23:33:39.307095Z","title":"Maybe only 0.5\\ arXiv preprint arXiv:2305.09246, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.06524","last_updated":"2025-09-08T10:30:58Z","snapshot_observed_at":"2026-08-07T01:24:37.601184Z","submitted_at":"2025-09-08T10:30:58Z","title":"LAMDAS: LLM as an Implicit Classifier for Domain-specific Data Selection","version":1},"reference_index":5,"source":"arxiv_source","source_observed_at":"2026-08-04T23:33:39.307095Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2509.06524"},"observation_digest":"sha256:09ef0c42b275ccec99325a93abc1e1a302df82ae49eca94f6ece8f92fe95dc17","observation_id":"31f8ba62-0d74-4f2a-93fa-901086584fb9","resolution":{"observed_at":"2026-08-04T23:33:39.307095Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-04T20:55:41.527188Z","title":"Maybe only 0.5% data is needed: A preliminary exploration of low training data instruction tuning,","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2509.08269","last_updated":"2026-07-17T08:53:46Z","snapshot_observed_at":"2026-08-08T15:31:01.848037Z","submitted_at":"2025-09-10T04:05:54Z","title":"A Systematic Survey on Large Language Models for Evolutionary Optimization: From Modeling to Solving","version":6},"reference_index":110,"source":"pdf_text","source_observed_at":"2026-08-04T20:55:41.527188Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2509.08269"},"observation_digest":"sha256:3ed6ba495952b46c39a984d943e843fd988b38d34ccbdf1af3bd8e6566643c86","observation_id":"d40b9ff7-d536-4acf-9353-ca8e723307c4","resolution":{"observed_at":"2026-08-04T20:55:41.527188Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-02T23:10:59.419614Z","title":"Maybe only 0.5\\ ArXiv preprint, abs/2305.09246, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2602.14696","last_updated":"2026-06-18T11:23:03Z","snapshot_observed_at":"2026-08-09T03:31:36.084800Z","submitted_at":"2026-02-16T12:33:05Z","title":"A Critical Look at Targeted Instruction Selection: Disentangling What Matters (and What Doesn't)","version":2},"reference_index":14,"source":"arxiv_source","source_observed_at":"2026-08-02T23:10:59.419614Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2602.14696"},"observation_digest":"sha256:3cc7a97423f353e1c4570820a8b346455e80e69f21ea60b49dc2fe26ec586b13","observation_id":"0e20ff72-985d-4a07-8e15-ce9c79849849","resolution":{"observed_at":"2026-08-02T23:10:59.419614Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":"2305.09246","doi":"10.48550/arxiv.2305.09246","metadata_source":"arxiv_reference","pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2305.09246 , year=","venue":"arXiv (Cornell University)","work_id":"8815fb68-5998-4c8e-aa14-382ab7927214","year":2023},"citing_paper":{"arxiv_id":"2604.11810","last_updated":"2026-04-09T14:08:01Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2026-04-09T14:08:01Z","title":"GRACE: A Dynamic Coreset Selection Framework for Large Language Model Optimization","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-05-10T18:06:46.131725Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2604.11810"},"observation_digest":"sha256:a833263999b6bd0cb78a0c237b37661625af4d47c399866ff77e93b1093dde3b","observation_id":"3dc571b7-546d-4855-b965-92104cab70fc","resolution":{"observed_at":"2026-05-10T20:30:48.972081Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":"2305.09246","doi":"10.48550/arxiv.2305.09246","metadata_source":"arxiv_reference","pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2305.09246 , year=","venue":"arXiv (Cornell University)","work_id":"8815fb68-5998-4c8e-aa14-382ab7927214","year":2023},"citing_paper":{"arxiv_id":"2604.17650","last_updated":"2026-04-19T22:45:10Z","snapshot_observed_at":"2026-07-06T23:04:41.564912Z","submitted_at":"2026-04-19T22:45:10Z","title":"Measuring Distribution Shift in User Prompts and Its Effects on LLM Performance","version":1},"reference_index":71,"source":"arxiv_source","source_observed_at":"2026-05-10T05:27:50.049798Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2604.17650"},"observation_digest":"sha256:4be70aa7a4c5e350084f0b41680dddd861eb65ad5a2fe4886c70fdcc7fb19d1a","observation_id":"0d257c10-9733-40ce-8402-a1a36c6abb3e","resolution":{"observed_at":"2026-05-10T06:56:47.928366Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":"2305.09246","doi":"10.48550/arxiv.2305.09246","metadata_source":"arxiv_reference","pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2305.09246 , year=","venue":"arXiv (Cornell University)","work_id":"8815fb68-5998-4c8e-aa14-382ab7927214","year":2023},"citing_paper":{"arxiv_id":"2605.00369","last_updated":"2026-06-05T01:05:58Z","snapshot_observed_at":"2026-07-06T23:13:47.082550Z","submitted_at":"2026-05-01T03:12:16Z","title":"InvEvolve: Evolving White-Box Inventory Policies via Large Language Models with Performance Guarantees","version":2},"reference_index":166,"source":"arxiv_source","source_observed_at":"2026-05-09T19:50:39.734124Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2605.00369"},"observation_digest":"sha256:2e9d35f1f07ad667f4aa5518e66d8430dbee9a12eb33e104fd869798800e6d9e","observation_id":"b8741f3e-86b9-44f7-a692-e87c6dbefe66","resolution":{"observed_at":"2026-05-11T15:31:08.034102Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning","version":1},"cited_work":{"arxiv_id":"2305.09246","doi":"10.48550/arxiv.2305.09246","metadata_source":"arxiv_reference","pith_arxiv_id":"2305.09246","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2305.09246 , year=","venue":"arXiv (Cornell University)","work_id":"8815fb68-5998-4c8e-aa14-382ab7927214","year":2023},"citing_paper":{"arxiv_id":"2605.00369","last_updated":"2026-06-05T01:05:58Z","snapshot_observed_at":"2026-07-06T23:13:47.082550Z","submitted_at":"2026-05-01T03:12:16Z","title":"InvEvolve: Evolving White-Box Inventory Policies via Large Language Models with Performance Guarantees","version":3},"reference_index":166,"source":"arxiv_source","source_observed_at":"2026-05-12T02:38:45.322351Z"},"links":{"cited_paper":"/paper/2305.09246","citing_paper":"/paper/2605.00369"},"observation_digest":"sha256:00825468fd51ee0806191e97dc827e6704fb2fdcb2a58f6def563f29c5c035dc","observation_id":"f5bd5196-42f2-4383-a605-f82d92ada5f5","resolution":{"observed_at":"2026-05-12T02:41:17.650267Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"state":"measured"}}],"links":{"evidence":"/evidence","html":"/paper/2305.09246/citation-record","integrity":"/paper/2305.09246/integrity","json":"/paper/2305.09246/citation-record.json","paper":"/paper/2305.09246"},"outbound":[],"paper":{"arxiv_id":"2305.09246","last_updated":"2023-05-16T07:52:57Z","latest_version":1,"primary_category":"cs.AI","snapshot_observed_at":"2026-08-05T06:19:25.316529Z","submitted_at":"2023-05-16T07:52:57Z","title":"Maybe Only 0.5% Data is Needed: A Preliminary Exploration of Low Training Data Instruction Tuning"},"reference_resolution":{"displayed":0,"state_counts":{"malformed_identifier":0,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":0,"verified_exact":0,"verified_fuzzy":0},"total_outbound_references":0},"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":"2026-08-09T06:31:02.800959+00:00","source":"crossref"},{"observed_at":"2026-08-09T06:30:57.326959+00:00","source":"retraction_watch"}],"thesis":"As of 10 August 2026, this Paper Citation Record lists 0 of 0 outbound references and 14 inbound Pith citation observations for arXiv:2305.09246."}