{"as_of":"2026-08-07T02:28:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:ebefa6b3476d2ac6f6fa2c8cefdcc573a7b2c5c4e0ea3989c56b9ee0bf9c8e57","coverage":[{"denominator":60,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":60,"source":"paper_references, paper_reference_links","source_observed_at":"2026-05-12T03:58:11.179607Z","state":"measured"},{"denominator":60,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":60,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-06T06:34:29.942622+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2605.09995/citation-record","integrity":"/paper/2605.09995/integrity","json":"/paper/2605.09995/citation-record.json","paper":"/paper/2605.09995"},"outbound":[{"citation":{"cited_paper":{"arxiv_id":"2504.05228","last_updated":"2025-08-09T18:33:13Z","snapshot_observed_at":"2026-08-01T15:38:55.010903Z","submitted_at":"2025-04-07T16:14:23Z","title":"NoveltyBench: Evaluating Language Models for Humanlike Diversity","version":4},"cited_work":{"arxiv_id":"2504.05228","doi":"10.48550/arxiv.2504.05228","metadata_source":"arxiv_reference","pith_arxiv_id":"2504.05228","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Noveltybench: Evaluating language models for humanlike diver- sity","venue":"ArXiv.org","work_id":"aba32c56-61d8-4972-9c8b-b852f95b11b0","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":1,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2504.05228","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:20fe46e5e1d2d1cdb8b28331c4fa5979754692b7379f854d7cebf5d7944bcdde","observation_id":"6f5d2ecf-c661-4bff-ba64-02597be182f6","resolution":{"observed_at":"2026-05-12T06:46:51.414237Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2510.01171","last_updated":"2026-07-15T22:07:56Z","snapshot_observed_at":"2026-08-04T13:23:08.839451Z","submitted_at":"2025-10-01T17:55:37Z","title":"Verbalized Sampling: How to Mitigate Mode Collapse and Unlock LLM Diversity","version":4},"cited_work":{"arxiv_id":"2510.01171","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2510.01171","snapshot_observed_at":"2026-07-17T01:20:39.022855Z","title":"Tomz, Christopher D","venue":null,"work_id":"ba11dce9-31c2-4b87-a872-45b7102fa956","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":2,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2510.01171","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:2d16d642d0a39f34bfc724fea8eb992f2543b2ff6888e1175c6e93c787bb1b24","observation_id":"91b0cc4c-b9e0-43f7-a4d3-fb1d5577af8d","resolution":{"observed_at":"2026-07-17T01:20:39.022855Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2310.06452","last_updated":"2024-02-19T14:39:07Z","snapshot_observed_at":"2026-07-29T18:39:02.141391Z","submitted_at":"2023-10-10T09:25:44Z","title":"Understanding the Effects of RLHF on LLM Generalisation and Diversity","version":3},"cited_work":{"arxiv_id":"2310.06452","doi":"10.48550/arxiv.2310.06452","metadata_source":"pith","pith_arxiv_id":"2310.06452","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Understanding the Effects of RLHF on LLM Generalisation and Diversity","venue":"cs.LG","work_id":"13d47639-3500-414f-b6ea-1f277577ad3b","year":2023},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":3,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2310.06452","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:a50d825a210ac3a8c90cf0c33fa5f0b600e76cf74992fcee87470627ac5cb220","observation_id":"c87b1bec-0986-4b12-8a16-49e91f333a20","resolution":{"observed_at":"2026-05-19T02:34:44.407176Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.18949","last_updated":"2025-05-25T02:52:35Z","snapshot_observed_at":"2026-07-06T21:30:00.261435Z","submitted_at":"2025-05-25T02:52:35Z","title":"The Price of Format: Diversity Collapse in LLMs","version":1},"cited_work":{"arxiv_id":"2505.18949","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2505.18949","snapshot_observed_at":"2026-06-29T18:13:48.857077Z","title":"The price of format: Diversity collapse in llms","venue":null,"work_id":"855c38a3-27e1-46dc-8563-8b8cef8a95b0","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":4,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2505.18949","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:3dca8f7608ec31781ad9f5a951d4209804d26b03c799c5d1b5f6af8997280023","observation_id":"7dbb3f32-f174-4b92-8ee4-b80ed4753cae","resolution":{"observed_at":"2026-05-12T06:46:51.363060Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":null,"venue":null,"work_id":"5c11fa45-0935-4de9-8380-ce650c1d73c8","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":5,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:5082bac6c755b74e0e276fb1caba6e008d4e241721e0cd5f238329c57e1dceb4","observation_id":"7d73aed4-d71d-47e7-a449-f84a75797380","resolution":{"observed_at":"2026-05-12T17:21:44.116151Z","resolver_source":"raw_fallback","status":"unresolved"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2402.04477","last_updated":"2024-02-06T23:52:58Z","snapshot_observed_at":"2026-08-06T05:04:14.520030Z","submitted_at":"2024-02-06T23:52:58Z","title":"Detecting Mode Collapse in Language Models via Narration","version":1},"cited_work":{"arxiv_id":"2402.04477","doi":null,"metadata_source":"pith","pith_arxiv_id":"2402.04477","snapshot_observed_at":"2026-07-10T13:47:05.741069Z","title":"arXiv preprint arXiv:2402.04477 , year=","venue":"cs.CL","work_id":"0b9387e6-8c25-4177-a217-deea1655c84c","year":2024},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":6,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2402.04477","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:6ed3660f98ceda74dc3af6a0445a56139f287629eddaaf17a22a4137dcd64451","observation_id":"fec62708-8aa5-4f38-aba1-cdb220115bfc","resolution":{"observed_at":"2026-05-12T06:46:51.291155Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Transactions of the Association for Computational Linguistics , volume=","venue":null,"work_id":"41860fd9-c297-435d-bb78-b3cf02db5898","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":7,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:ff9cb7e130cf3b7dc76fa952aa31f8cdebc7239057ab3e65002b4f3419b7269d","observation_id":"e0dcf735-46d4-4447-9db8-38cfdb2e5f39","resolution":{"observed_at":"2026-05-12T17:21:44.132627Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2408.16673","last_updated":"2025-04-05T08:56:18Z","snapshot_observed_at":"2026-07-06T19:07:46.545514Z","submitted_at":"2024-08-29T16:21:00Z","title":"Preserving Diversity in Supervised Fine-Tuning of Large Language Models","version":2},"cited_work":{"arxiv_id":"2408.16673","doi":"10.48550/arxiv.2408.16673","metadata_source":"arxiv_reference","pith_arxiv_id":"2408.16673","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Preserving diversity in supervised fine-tuning of large language models","venue":"arXiv (Cornell University)","work_id":"161bcac5-10cd-4c59-8075-24b77ef37471","year":2024},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":8,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2408.16673","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:c7796c0398dd3013f8fe97f2be71dd14796eecd312de6d4600ff8e5a141aa107","observation_id":"cbcddc21-6c03-4b60-9fd0-4b03b4ddcf26","resolution":{"observed_at":"2026-05-12T06:46:51.299112Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2404.10859","last_updated":"2024-08-07T21:57:20Z","snapshot_observed_at":"2026-07-06T18:01:10.003838Z","submitted_at":"2024-04-16T19:17:23Z","title":"Forcing Diffuse Distributions out of Language Models","version":2},"cited_work":{"arxiv_id":"2404.10859","doi":"10.48550/arxiv.2404.10859","metadata_source":"arxiv_reference","pith_arxiv_id":"2404.10859","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Forcing diffuse distributions out of language models","venue":"arXiv (Cornell University)","work_id":"8b576a88-3267-4770-aa8c-3c92d4910dd5","year":2024},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":9,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2404.10859","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:1daf1035c943685a1218aa0393102a2c63a171f584d1014df730f9a52cdadc33","observation_id":"0efb7e35-540d-4765-be62-f1295efaa77d","resolution":{"observed_at":"2026-05-12T06:46:51.341849Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2511.05650","last_updated":"2026-06-01T17:26:03Z","snapshot_observed_at":"2026-08-03T23:32:50.566210Z","submitted_at":"2025-11-07T19:00:01Z","title":"Optimizing Diversity and Quality through Base-Aligned Model Collaboration","version":2},"cited_work":{"arxiv_id":"2511.05650","doi":null,"metadata_source":"pith","pith_arxiv_id":"2511.05650","snapshot_observed_at":"2026-06-29T07:33:13.843237Z","title":"arXiv preprint arXiv:2511.05650 , year=","venue":"cs.CL","work_id":"5d736cc6-0ee0-4485-9e4e-159c2196d79a","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":10,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2511.05650","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:ee8ea229b05f9efbdb49c62e8d0fab139645b81bbbf1519d47d8c44c0cefa7d0","observation_id":"dad48712-cd04-4733-ad60-c2cd3a3851a1","resolution":{"observed_at":"2026-06-02T04:04:48.600033Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-09T17:56:26.096579Z","title":"Advances in neural information processing systems , volume=","venue":null,"work_id":"c25e8154-fab2-455c-8a26-56e40aed5d2b","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":11,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:f30439a05b0b259299ee1b7e838661749d6e1448f7e0675ed215be29461387de","observation_id":"f526129b-24a6-46fd-86a4-cfd0072fc113","resolution":{"observed_at":"2026-05-12T17:21:44.125148Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1909.08593","last_updated":"2020-01-08T23:02:36Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2019-09-18T17:33:39Z","title":"Fine-Tuning Language Models from Human Preferences","version":2},"cited_work":{"arxiv_id":"1909.08593","doi":"10.48550/arxiv.1909.08593","metadata_source":"pith","pith_arxiv_id":"1909.08593","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Fine-Tuning Language Models from Human Preferences","venue":"cs.CL","work_id":"4f54aad1-f3b6-404f-b9c7-e21ba0a33b99","year":2019},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":12,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1909.08593","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:9655fe1a46f8fa0cd59393b23db77976b86a161550305348caceb8ff2d95ec24","observation_id":"c83977a8-18e0-4a17-b4f3-54790d77b534","resolution":{"observed_at":"2026-05-12T06:46:52.269644Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-07-18T08:21:06.631932+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-18T08:21:06.631932+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-11T02:47:56.151399Z","title":"Advances in neural information processing systems , volume=","venue":null,"work_id":"f69d9f4f-b804-4c65-835e-f340310c7748","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":13,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:016d82f27eec219fbdfd02019a08c843ffe2cc05bc891078b413694edb5830b1","observation_id":"a203194c-2336-45b2-959e-d53eb5c6ecc9","resolution":{"observed_at":"2026-05-12T17:21:44.176288Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"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":"2204.05862","doi":"10.1016/j.respol.2005.01.014","metadata_source":"pith","pith_arxiv_id":"2204.05862","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback","venue":"cs.CL","work_id":"a1f2574b-a899-4713-be60-c87ba332656c","year":2022},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":14,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2204.05862","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:f4ea3cfec65f249a617f34bbdea13cd17aa2e8768bdf1f93978fd90e031eceed","observation_id":"ec871030-1a54-423b-b56e-f06a678a4c04","resolution":{"observed_at":"2026-05-12T06:46:52.047015Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-11T02:47:55.730920Z","title":"Advances in neural information processing systems , volume=","venue":null,"work_id":"b0092220-e738-45d3-97b1-bd1f0bd834c5","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":15,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:2f72cdc72aeaff19afe7686496dc5a57ef266a44733565e1234b265f66c1b06c","observation_id":"c4879186-c37a-4d42-b6ec-f025fc2a9da8","resolution":{"observed_at":"2026-05-12T17:21:44.180577Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1908.04319","last_updated":"2019-09-26T23:57:44Z","snapshot_observed_at":"2026-08-05T17:26:35.920382Z","submitted_at":"2019-08-12T18:09:04Z","title":"Neural Text Generation with Unlikelihood Training","version":2},"cited_work":{"arxiv_id":"1908.04319","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"1908.04319","snapshot_observed_at":"2026-07-04T19:40:07.174377Z","title":"Neural text generation with unlikelihood training","venue":null,"work_id":"f5c0999f-3f2c-4fd2-be5b-486e99399ccf","year":2020},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":16,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1908.04319","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:79f10c4e92d4ab9237397db1d79b45363341d5c9fe0680131f417606db23a1b2","observation_id":"5a6b95d2-3897-4e3a-b717-2d226bb7c158","resolution":{"observed_at":"2026-05-12T06:46:51.960967Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-05T07:30:47.442654Z","title":"Proceedings of the 2016 conference of the North American chapter of the association for computational linguistics: human language technologies , pages=","venue":null,"work_id":"e5ebb387-df73-4b25-9120-3c560b22ceee","year":2016},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":17,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:23369f208df6e6035d908f50008a5440812f8377eb73af48f3578141f7d58ba6","observation_id":"2a090592-befe-445f-a400-3a43ec6347ac","resolution":{"observed_at":"2026-05-12T17:21:44.129035Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1703.10960","last_updated":"2017-10-21T04:58:20Z","snapshot_observed_at":"2026-08-05T04:49:46.412328Z","submitted_at":"2017-03-31T15:55:00Z","title":"Learning Discourse-level Diversity for Neural Dialog Models using Conditional Variational Autoencoders","version":3},"cited_work":{"arxiv_id":"1703.10960","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"1703.10960","snapshot_observed_at":"2026-07-04T22:17:23.878014Z","title":"arXiv preprint arXiv:1703.10960 , year=","venue":null,"work_id":"ac644da0-45de-4825-976f-e77b228f2e0e","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":18,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1703.10960","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:2e1045acb6fc8d5113042780a0d0081b18c372d3f54d25df4b236b2cee79a8cf","observation_id":"e0c4141a-eda7-426b-98be-72c59ac02877","resolution":{"observed_at":"2026-07-04T22:17:23.878014Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Latent Variable Dialogue Models and their Diversity","venue":null,"work_id":"154208e8-1736-4d9b-a386-3c9e00aca3c2","year":2017},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":19,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:21e93cdc0ddcaac8cd813c0b4e04cd905d0cf8f3dd385d009e6c679de4d9a5a4","observation_id":"dbf8ba8d-9348-4938-ba95-429bfce2a813","resolution":{"observed_at":"2026-05-12T17:21:44.160768Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Proceedings of the AAAI conference on artificial intelligence , volume=","venue":null,"work_id":"bfb1ec88-234f-4718-9f0d-e942c0d265e5","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":20,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:9703a9471ae1dfbb4cf9e6edbb9aaca75195be3be2332c64dc11b2ed526c5dbe","observation_id":"bf1785ca-6d78-4d3a-b971-9a96717d6485","resolution":{"observed_at":"2026-05-12T17:21:44.189590Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2210.12326","last_updated":"2022-10-22T01:57:16Z","snapshot_observed_at":"2026-08-05T18:10:12.218921Z","submitted_at":"2022-10-22T01:57:16Z","title":"Transformer-Based Conditioned Variational Autoencoder for Dialogue Generation","version":1},"cited_work":{"arxiv_id":"2210.12326","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2210.12326","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"arXiv preprint arXiv:2210.12326 , year=","venue":null,"work_id":"b8127770-41a3-4b6b-8e98-bed17be0fac9","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":21,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2210.12326","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:b74516dcac6dd33989f109fe861c0188dabfebef6ff898a3a3fe311cee9839f7","observation_id":"e317bef0-e3d1-44dd-ad4d-e0f14d20a605","resolution":{"observed_at":"2026-05-12T06:46:52.218236Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1909.05858","last_updated":"2019-09-20T20:08:56Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2019-09-11T17:57:18Z","title":"CTRL: A Conditional Transformer Language Model for Controllable Generation","version":2},"cited_work":{"arxiv_id":"1909.05858","doi":null,"metadata_source":"pith","pith_arxiv_id":"1909.05858","snapshot_observed_at":"2026-07-10T11:37:03.420559Z","title":"CTRL: A Conditional Transformer Language Model for Controllable Generation","venue":"cs.CL","work_id":"c90b18cd-644d-4503-8eca-953249853cff","year":2019},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":22,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1909.05858","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:0f4daf8c590cd476544226bf3b03c4fc7b7b3a5827c0046a27c91ac74337c8aa","observation_id":"0b9bfa53-27a0-4b1a-97b2-07ec794c0172","resolution":{"observed_at":"2026-05-17T06:14:02.733439Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-05T04:40:41.455908Z","title":"Advances in Neural Information Processing Systems , volume=","venue":null,"work_id":"df82cbb3-4458-46ba-bf08-f8f4c1eccbd5","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":23,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:4b508e1e0a776c160629080b71076566b2a6c12e9445fc9b2eb89a24be5570ba","observation_id":"42756b15-6022-49de-8a99-be2cae62b679","resolution":{"observed_at":"2026-05-12T17:21:44.197566Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1904.09751","last_updated":"2020-02-14T21:56:30Z","snapshot_observed_at":"2026-07-06T07:47:32.745963Z","submitted_at":"2019-04-22T07:17:18Z","title":"The Curious Case of Neural Text Degeneration","version":2},"cited_work":{"arxiv_id":"1904.09751","doi":"10.48550/arxiv.1904.09751","metadata_source":"pith","pith_arxiv_id":"1904.09751","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"The Curious Case of Neural Text Degeneration","venue":"cs.CL","work_id":"1ef2ec4a-db7b-42b2-9416-0f1bb628c3c8","year":2019},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":24,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1904.09751","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:17cd5fd73c345a0d7f6ef5d8eacbed0d6ccf489e226db46f49bf8fdc6ea5f842","observation_id":"9aaa7e45-eef5-49df-86bd-5fd8b56837bb","resolution":{"observed_at":"2026-05-12T06:46:52.295530Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-05-19T15:54:15.356125+00:00","source":"crossref_status_cache"},{"observed_at":"2026-05-19T15:54:15.356125+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1811.02549","last_updated":"2020-02-19T22:44:37Z","snapshot_observed_at":"2026-08-04T12:55:07.027032Z","submitted_at":"2018-11-06T18:44:11Z","title":"Language GANs Falling Short","version":6},"cited_work":{"arxiv_id":"1811.02549","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"1811.02549","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Language GANs falling short","venue":null,"work_id":"851f20c8-1e5a-4d79-8bac-46b38d0a46aa","year":2018},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":25,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1811.02549","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:4661f67c2a283ef81ce961a76cbe3e8cbd227af36c22e0f6ec337cfd205865f6","observation_id":"7524a878-da00-43fe-97ba-366789a1b7b3","resolution":{"observed_at":"2026-05-12T06:46:51.948375Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1610.02424","last_updated":"2018-10-22T13:48:32Z","snapshot_observed_at":"2026-07-06T05:13:49.420787Z","submitted_at":"2016-10-07T20:56:47Z","title":"Diverse Beam Search: Decoding Diverse Solutions from Neural Sequence Models","version":2},"cited_work":{"arxiv_id":"1610.02424","doi":"10.48550/arxiv.1610.02424","metadata_source":"pith","pith_arxiv_id":"1610.02424","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Diverse Beam Search: Decoding Diverse Solutions from Neural Sequence Models","venue":"cs.AI","work_id":"e1f4742b-32a5-4228-a299-1993c7c90d7b","year":2016},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":26,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1610.02424","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:26fa04f397b7bfd9dad492008d791165c9aa8ce6c3169d8559fce51d1620805f","observation_id":"eed1aac0-2b4b-41ee-b45a-7234c971a794","resolution":{"observed_at":"2026-05-12T06:46:51.869367Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"International Conference on Machine Learning , pages=","venue":null,"work_id":"ba494d9f-9aa7-40ba-a721-76ba877661b8","year":2023},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":27,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:2daaf4440e5a3771b2fdfc8e13c280c30b66be0ed824b541a8a1fc2c3e2c6ff1","observation_id":"e1018603-ad80-4948-8822-ff8a435921e9","resolution":{"observed_at":"2026-05-12T17:21:44.153380Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.00047","last_updated":"2025-09-13T00:52:27Z","snapshot_observed_at":"2026-07-06T21:17:14.657775Z","submitted_at":"2025-04-30T03:41:55Z","title":"Base Models Beat Aligned Models at Randomness and Creativity","version":2},"cited_work":{"arxiv_id":"2505.00047","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2505.00047","snapshot_observed_at":"2026-07-03T05:47:41.530108Z","title":"Base models beat aligned models at random- ness and creativity","venue":null,"work_id":"f446c0cf-a5c8-4844-8e79-4b1d29e8337c","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":28,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2505.00047","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:c1d65306f93a87b0c4cd704ae4094e19ff69aca50a447b9cbde938f1eb409066","observation_id":"49a342f1-b9c3-4977-a900-f41c057ad8b7","resolution":{"observed_at":"2026-05-12T06:46:51.955719Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-09T15:16:18.971970Z","title":"Advances in neural information processing systems , volume=","venue":null,"work_id":"4a77e424-cf16-4c16-8d94-2ee44db893d1","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":29,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:3907a3508fb7ce6ab60e53a7ccc6182d5d103f50ff286d8650297187721accdb","observation_id":"f396f6aa-5247-4484-8721-4cacd38c5135","resolution":{"observed_at":"2026-05-12T17:21:44.209552Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1912.02164","last_updated":"2020-03-03T05:33:49Z","snapshot_observed_at":"2026-07-06T08:42:03.833704Z","submitted_at":"2019-12-04T18:32:15Z","title":"Plug and Play Language Models: A Simple Approach to Controlled Text Generation","version":4},"cited_work":{"arxiv_id":"1912.02164","doi":"10.48550/arxiv.1912.02164","metadata_source":"pith","pith_arxiv_id":"1912.02164","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Yosinski, and Rosanne Liu","venue":"cs.CL","work_id":"0b303bb5-76b4-444d-bc87-1da9c1673b7a","year":2019},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":30,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1912.02164","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:be579ce4b521e4335a4fd79b3babb591119704834e615b6dd6c03753d59e5d48","observation_id":"5bb04422-88ff-48dc-ad51-b29901f31ad7","resolution":{"observed_at":"2026-05-12T06:46:51.864948Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1805.04833","last_updated":"2018-05-13T07:07:08Z","snapshot_observed_at":"2026-07-06T06:38:48.758485Z","submitted_at":"2018-05-13T07:07:08Z","title":"Hierarchical Neural Story Generation","version":1},"cited_work":{"arxiv_id":"1805.04833","doi":"10.48550/arxiv.1805.04833","metadata_source":"pith","pith_arxiv_id":"1805.04833","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Hierarchical Neural Story Generation","venue":"cs.CL","work_id":"5e06c7f2-e43a-4c73-b898-2848e095ff14","year":2018},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":31,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1805.04833","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:25ab79d411da18366ceefe83adc8aeb609793b168dbd12f2930c5a87478b5182","observation_id":"1fcc5542-3610-42ec-a1ee-671d5a056261","resolution":{"observed_at":"2026-05-12T06:46:52.116544Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-05-25T06:23:41.809346+00:00","source":"crossref_status_cache"},{"observed_at":"2026-05-25T06:23:41.809346+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2309.05196","last_updated":"2024-07-01T16:36:30Z","snapshot_observed_at":"2026-07-06T16:16:45.463211Z","submitted_at":"2023-09-11T02:16:47Z","title":"Does Writing with Language Models Reduce Content Diversity?","version":3},"cited_work":{"arxiv_id":"2309.05196","doi":"10.48550/arxiv.2309.05196","metadata_source":"pith","pith_arxiv_id":"2309.05196","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2309.05196 , year=","venue":"cs.CL","work_id":"580a91ce-6c32-4965-bf5b-0c9dae67ad72","year":2023},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":32,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2309.05196","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:3cf3dd58823dd7e6d4f0168db7dd3a3c282fa0b0d07c16d0f6491f46395928db","observation_id":"d16f8425-7146-4195-8b40-a50581d428a9","resolution":{"observed_at":"2026-05-12T06:46:51.875587Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume , pages=","venue":null,"work_id":"2c941191-955c-407e-b41e-8ba809801bbe","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":33,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:b326e6bed01946923e3887d58f95a2363ba06ca9d22f07a3422e5a80492f6a17","observation_id":"d8f0f776-cef1-4ec1-b19c-b6ec82efe37a","resolution":{"observed_at":"2026-05-12T17:21:44.193584Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1904.09675","last_updated":"2020-02-24T18:59:28Z","snapshot_observed_at":"2026-07-29T15:42:51.774083Z","submitted_at":"2019-04-21T23:08:53Z","title":"BERTScore: Evaluating Text Generation with BERT","version":3},"cited_work":{"arxiv_id":"1904.09675","doi":"10.48550/arxiv.1904.09675","metadata_source":"pith","pith_arxiv_id":"1904.09675","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"BERTScore: Evaluating Text Generation with BERT","venue":"cs.CL","work_id":"9eaaaac1-0a96-4f5f-9b13-30c46e9e1346","year":2019},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":34,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1904.09675","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:0bebb8efc1705ebec2a773962d767e5fea2b341c3f4ecba509669bdd1e082e3b","observation_id":"571f0bf3-ba74-4243-808c-3b08a0657695","resolution":{"observed_at":"2026-05-12T06:46:51.330628Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-04T01:08:14.227491+00:00","source":"crossref_status_cache"},{"observed_at":"2026-08-04T01:08:14.227491+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"The 41st international ACM SIGIR conference on research & development in information retrieval , pages=","venue":null,"work_id":"2abe9ac9-f598-4846-837f-5fbb4cb7b942","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":35,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:9a1d800e2258076ad8f3697e442fe2560adde69a97ed16ce8c789a50b20086a3","observation_id":"de29cfc4-a2ab-4596-a21e-a0e464a11624","resolution":{"observed_at":"2026-05-12T17:21:44.149561Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"ICLR 2024 Workshop on Mathematical and Empirical Understanding of Foundation Models , volume=","venue":null,"work_id":"b8599e83-b382-48d4-98bc-849c66ee6451","year":2024},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":36,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:80e499f3ec9c24b116a97b6a9eb34934870de8f320e39a2212b678d7590a8dbb","observation_id":"3d65616e-d9be-4fb7-9cfd-fba499a0220d","resolution":{"observed_at":"2026-05-12T17:21:44.157184Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"1701.06548","last_updated":"2017-01-23T18:35:28Z","snapshot_observed_at":"2026-08-05T00:39:48.314299Z","submitted_at":"2017-01-23T18:35:28Z","title":"Regularizing Neural Networks by Penalizing Confident Output Distributions","version":1},"cited_work":{"arxiv_id":"1701.06548","doi":null,"metadata_source":"pith","pith_arxiv_id":"1701.06548","snapshot_observed_at":"2026-07-01T12:15:43.664113Z","title":"Regularizing Neural Networks by Penalizing Confident Output Distributions","venue":"cs.NE","work_id":"bf8e939e-9c4a-41d0-8707-c7ca211c121a","year":2017},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":37,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/1701.06548","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:4733fd785b7bf6da73a498874b2a8c9218265787ad6c5d1f4715e846c4377afd","observation_id":"0e3b5849-a284-49d8-b0b0-6e5853af632d","resolution":{"observed_at":"2026-05-12T06:46:51.750366Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"International conference on machine learning , pages=","venue":null,"work_id":"8df4ab13-4fb4-445e-8265-3e529ec729d9","year":2019},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":38,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:7f8ccbeb06771e7670b2c6cd8041ec349bee6a2523c2ff0d97d9edcebe55c21c","observation_id":"8a73fa0e-6006-459b-b04d-5c79ef0d3f62","resolution":{"observed_at":"2026-05-12T17:21:44.145297Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Transactions of the Association for Computational Linguistics , volume=","venue":null,"work_id":"6cebc97f-b801-42bd-919f-4aae7bb7a975","year":2023},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":39,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:fc77962d0e1c9caddeaab3ffb497566fa737bea756270ede1e42e3f6fb08ab99","observation_id":"b0b9cec7-8a74-4771-ac76-45616ccdb3a3","resolution":{"observed_at":"2026-05-12T17:21:44.202291Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Advances in Neural Information Processing Systems , volume=","venue":null,"work_id":"a5745d4f-e1d6-461e-8bbf-fe64716473b8","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":40,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:7c508e8c2afd03b7de6d88a4b9c506f79056c2eb034732b4b7f72f8332211443","observation_id":"ce077972-2e30-48df-9bb7-44c52daea008","resolution":{"observed_at":"2026-05-12T17:21:44.141378Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Proceedings of the AAAI Conference on Artificial Intelligence , volume=","venue":null,"work_id":"1ebeeeee-801e-48d8-b183-01dd1a1bcc52","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":41,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:bc959a770a849e39773cdc45cb20b435e56ef905a1eaf5f50a53cfa6cbb3c061","observation_id":"90ff7f0f-df60-4ac9-8bf0-a7c745afbbce","resolution":{"observed_at":"2026-05-12T17:21:44.164298Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-09T17:56:26.061201Z","title":"Proceedings of the national academy of sciences , volume=","venue":null,"work_id":"47e891d1-91f8-4e1a-9923-74473e0b4b20","year":2017},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":42,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:854baf869fc191b4432b34afae18059d34a70f3e160e9a7b4b58f9361f554b7f","observation_id":"374c9808-84a4-48cc-9e69-06f200106880","resolution":{"observed_at":"2026-05-12T17:21:44.168016Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Advances in Neural Information Processing Systems , volume=","venue":null,"work_id":"d4c1e1c9-4ee8-4ffd-8760-1fa9861f2918","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":43,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:4a3c16d676458d4b52a80c1142e1ae8419a0136c6870ec22a12100a3eeda9a3d","observation_id":"b3d9ac2d-28ca-489a-b445-15cf29e29466","resolution":{"observed_at":"2026-05-12T17:21:44.171836Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-10T00:26:39.962022Z","title":"IEEE transactions on pattern analysis and machine intelligence , volume=","venue":null,"work_id":"ee1babc2-e8f3-44d5-9709-865caf240a58","year":2017},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":44,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:42d3adbae35a3311ebd83e5f93beade3fd73d6fc9763ff917bec21fd20287f64","observation_id":"76ad3c07-14df-43e2-9175-9388a014447a","resolution":{"observed_at":"2026-05-12T17:21:44.120709Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Proceedings of the IEEE conference on Computer Vision and Pattern Recognition , pages=","venue":null,"work_id":"f0c9a772-553f-468f-91ee-2c10e197716f","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":45,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:a861b843b7968347b583fdf6ce2f21fa897fccfa4ceaec088d1be6dcdd5feadd","observation_id":"aebdaec0-8947-4f13-ad98-90262ebb3ccd","resolution":{"observed_at":"2026-05-12T17:21:44.137177Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-07-06T19:52:52.389735Z","title":"Advances in neural information processing systems , volume=","venue":null,"work_id":"a0351305-cc3b-49bc-a4b9-c753c124404e","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":46,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:4fed17707767f5d8730dbab308d4084ece323c8fe1f8e42c48ee63eeb6c2e2d3","observation_id":"1bfcb226-d036-4cd0-a075-455487991907","resolution":{"observed_at":"2026-05-12T17:21:44.108921Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2305.17493","last_updated":"2024-04-14T05:20:10Z","snapshot_observed_at":"2026-08-05T16:03:40.517679Z","submitted_at":"2023-05-27T15:10:41Z","title":"The Curse of Recursion: Training on Generated Data Makes Models Forget","version":3},"cited_work":{"arxiv_id":"2305.17493","doi":"10.48550/arxiv.2305.17493","metadata_source":"pith","pith_arxiv_id":"2305.17493","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"The Curse of Recursion: Training on Generated Data Makes Models Forget","venue":"cs.LG","work_id":"14bc9b3b-d8dd-4c25-a0be-6c178d31be0e","year":2023},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":47,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2305.17493","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:dc9f7926bd252eabf829cf63fc4f6a2ead1c03b50ac59ceb0ab2891cc6b589dd","observation_id":"7443ee2e-8b47-4b7b-8f89-3a299ecb3dfa","resolution":{"observed_at":"2026-05-20T14:04:58.330652Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-05-20T14:22:09.312587+00:00","source":"crossref_status_cache"},{"observed_at":"2026-05-20T14:22:09.312587+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"International conference on machine learning , pages=","venue":null,"work_id":"12f958c8-97a2-4701-a1b4-e33179ed971c","year":2018},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":48,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:ba81541c10ca573b034db51c61aefb3a121836d8882f7b4c5c9148b7eb2701f4","observation_id":"c5e8a5ec-8212-409d-afb9-ef36985283ed","resolution":{"observed_at":"2026-05-12T17:21:44.112589Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":"2510.22954","doi":"10.48550/arxiv.2510.22954","metadata_source":"arxiv_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"arXiv preprint arXiv:2510.22954 , year=","venue":"arXiv (Cornell University)","work_id":"c44a4fe8-40d7-423c-90c2-07786de43600","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":49,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:bd389089ddf17d01f99f3400f7cceefc2f4ac74acdd0899349a0588e3ce5133b","observation_id":"b2022dca-699c-4f1a-b213-26e0ec9ff4d4","resolution":{"observed_at":"2026-05-12T06:46:51.390764Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-05-23T16:24:49.91549+00:00","source":"crossref_status_cache"},{"observed_at":"2026-05-23T16:24:49.91549+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2506.09659","last_updated":"2025-06-11T12:26:45Z","snapshot_observed_at":"2026-07-06T21:40:21.542390Z","submitted_at":"2025-06-11T12:26:45Z","title":"Intent Factored Generation: Unleashing the Diversity in Your Language Model","version":1},"cited_work":{"arxiv_id":"2506.09659","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2506.09659","snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"arXiv preprint arXiv:2506.09659 , year=","venue":null,"work_id":"588a979b-5dd4-4e94-b5f6-dbadd7237319","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":50,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2506.09659","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:c2093973c8b64451bfcde3279d22a05389460e0a33241074bcd4ad89ce0495ed","observation_id":"fb4d8dff-0231-4ff9-a22a-f56c0a76e324","resolution":{"observed_at":"2026-05-12T06:46:51.368422Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2505.09388","last_updated":"2025-05-14T13:41:34Z","snapshot_observed_at":"2026-07-06T02:11:23.670680Z","submitted_at":"2025-05-14T13:41:34Z","title":"Qwen3 Technical Report","version":1},"cited_work":{"arxiv_id":"2505.09388","doi":"10.1016/j.aiopen.2022.12","metadata_source":"pith","pith_arxiv_id":"2505.09388","snapshot_observed_at":"2026-07-11T11:50:26.030339Z","title":"Qwen3 Technical Report","venue":"cs.CL","work_id":"25a4e30c-1232-48e7-9925-02fa12ba7c9e","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":51,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2505.09388","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:9237a8c335af8102bfd962b9ebb77074bc1772b4da27d6c2afd0b2e9cd2e6fa7","observation_id":"c74a5244-6d98-493e-805d-ec34e53efab6","resolution":{"observed_at":"2026-05-12T06:46:51.626888Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2407.21783","last_updated":"2024-11-23T23:27:33Z","snapshot_observed_at":"2026-07-06T18:55:11.576666Z","submitted_at":"2024-07-31T17:54:27Z","title":"The Llama 3 Herd of Models","version":3},"cited_work":{"arxiv_id":"2407.21783","doi":"10.1016/s0749-0720(15","metadata_source":"pith","pith_arxiv_id":"2407.21783","snapshot_observed_at":"2026-07-11T11:50:26.030339Z","title":"The Llama 3 Herd of Models","venue":"cs.AI","work_id":"1549a635-88af-4ac1-acfe-51ae7bb53345","year":2024},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":52,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2407.21783","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:7e6b2bcee6572ba87d308423aa46cea3ea52110991d1f0f572f49785ad03bbb9","observation_id":"93260900-2f15-4b93-8f5e-6d29ad300624","resolution":{"observed_at":"2026-05-12T06:46:51.373669Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2302.09664","last_updated":"2023-04-15T12:55:45Z","snapshot_observed_at":"2026-07-06T14:53:27.667483Z","submitted_at":"2023-02-19T20:10:07Z","title":"Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generation","version":3},"cited_work":{"arxiv_id":"2302.09664","doi":"10.48550/arxiv.2302.09664","metadata_source":"pith","pith_arxiv_id":"2302.09664","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generation","venue":"cs.CL","work_id":"d66d411b-c2c1-4cd6-8a6b-9fac872fa257","year":2023},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":53,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2302.09664","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:b7156a22188abb2c1398a89159e250bcdf29f7788bf31ea78f5c489b77cd91ab","observation_id":"311a6164-0ff7-45ad-9adb-1789baba2b49","resolution":{"observed_at":"2026-05-12T18:00:05.468379Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":"raw_reference","pith_arxiv_id":null,"snapshot_observed_at":"2026-06-05T21:23:00.469572Z","title":"Peters and Abhilasha Ravichander and Kyle Richardson and Zejiang Shen and Emma Strubell and Nishant Subramani and Oyvind Tafjord and Pete Walsh and Luke Zettlemoyer and Noah A","venue":null,"work_id":"148a795d-b6f5-48e7-a348-4a2de16254ac","year":null},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":54,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:6271445c22c37a0768e1cea5fbbbc38b9582f0528df4008441a1e56cda2593a7","observation_id":"a7e30364-b704-433f-a7e0-034b712c0c81","resolution":{"observed_at":"2026-05-12T17:21:44.184916Z","resolver_source":"raw_fallback","status":"verified_fuzzy"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2501.00656","last_updated":"2025-10-08T07:50:45Z","snapshot_observed_at":"2026-07-06T20:15:22.511263Z","submitted_at":"2024-12-31T21:55:10Z","title":"2 OLMo 2 Furious","version":3},"cited_work":{"arxiv_id":"2501.00656","doi":"10.48550/arxiv.2501.00656","metadata_source":"pith","pith_arxiv_id":"2501.00656","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"2 OLMo 2 Furious","venue":"cs.CL","work_id":"9ef0dc2b-fdfe-4f14-b235-ef7556dc709a","year":2024},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":55,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2501.00656","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:af6ad11f7db9772145e2890ddacedef925799d143cff91f56c48d1aac7defe47","observation_id":"1818f3a9-e391-4c33-b70e-f43be24c9589","resolution":{"observed_at":"2026-05-12T06:46:51.350104Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-07-11T05:19:24.20254+00:00","source":"crossref_status_cache"},{"observed_at":"2026-07-11T05:19:24.20254+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2203.15556","last_updated":"2022-03-29T13:38:03Z","snapshot_observed_at":"2026-07-06T12:54:11.616335Z","submitted_at":"2022-03-29T13:38:03Z","title":"Training Compute-Optimal Large Language Models","version":1},"cited_work":{"arxiv_id":"2203.15556","doi":"10.1098/rsta.2024.0522","metadata_source":"pith","pith_arxiv_id":"2203.15556","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Training Compute-Optimal Large Language Models","venue":"cs.CL","work_id":"b2faf28d-86b7-429c-bc42-469458efc246","year":2022},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":56,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2203.15556","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:5dcab9d2f18b68c40f6e3aa1a3440fcf9ece97c9f078c13aa80a1b5478d1e4b8","observation_id":"d6ba83a9-da25-4c86-996d-cfab2a9f58f7","resolution":{"observed_at":"2026-05-12T06:46:51.305020Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2411.15124","last_updated":"2025-04-14T22:39:09Z","snapshot_observed_at":"2026-07-06T19:55:37.400185Z","submitted_at":"2024-11-22T18:44:04Z","title":"Tulu 3: Pushing Frontiers in Open Language Model Post-Training","version":5},"cited_work":{"arxiv_id":"2411.15124","doi":"10.48550/arxiv.2411.15124","metadata_source":"pith","pith_arxiv_id":"2411.15124","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Tulu 3: Pushing Frontiers in Open Language Model Post-Training","venue":"cs.CL","work_id":"28c9dbea-056a-48c2-8000-85f809827e45","year":2024},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":57,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2411.15124","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:4a00c508da1d70c54d9277e2eac63469ae0a529defcd161ef2db0da4c507b5ca","observation_id":"9306fe6e-2703-4767-9758-2443bae75798","resolution":{"observed_at":"2026-05-12T06:46:51.913896Z","resolver_source":"local_arxiv","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-05-23T21:53:00.522112+00:00","source":"crossref_status_cache"},{"observed_at":"2026-05-23T21:53:00.522112+00:00","source":"openalex_status_cache"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2412.02674","last_updated":"2025-02-25T16:59:11Z","snapshot_observed_at":"2026-08-06T16:22:55.310872Z","submitted_at":"2024-12-03T18:47:26Z","title":"Mind the Gap: Examining the Self-Improvement Capabilities of Large Language Models","version":2},"cited_work":{"arxiv_id":"2412.02674","doi":null,"metadata_source":"arxiv_reference","pith_arxiv_id":"2412.02674","snapshot_observed_at":"2026-07-01T20:56:13.476979Z","title":"Mind the gap: Examining the self-improvement capabilities of large language models","venue":null,"work_id":"13db825e-1742-4503-aff5-6755fdc61624","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":58,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2412.02674","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:d7d2fcaac8446bbcf478a6d7c4a291e375514bfcc912ae2cb32fde168fe7a0ef","observation_id":"caec8726-3d6a-4f00-a7cb-0274bf57bdcc","resolution":{"observed_at":"2026-05-12T06:46:51.283999Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2405.01470","last_updated":"2024-05-02T17:00:02Z","snapshot_observed_at":"2026-07-06T18:08:56.480769Z","submitted_at":"2024-05-02T17:00:02Z","title":"WildChat: 1M ChatGPT Interaction Logs in the Wild","version":1},"cited_work":{"arxiv_id":"2405.01470","doi":"10.48550/arxiv.2405.01470","metadata_source":"pith","pith_arxiv_id":"2405.01470","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"WildChat: 1M ChatGPT Interaction Logs in the Wild","venue":"cs.CL","work_id":"799d0d7b-7d66-40bb-b17f-205e5d2f3e13","year":2024},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":59,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2405.01470","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:15da7fa069d8181e952b93be729892ee64293cba6889c314bcd67a18a4110088","observation_id":"64ea2ffb-3d3a-467b-afda-30b2cad0825e","resolution":{"observed_at":"2026-05-16T10:43:02.331896Z","resolver_source":"arxiv_id","status":"metadata_mismatch"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":{"arxiv_id":"2504.09184","last_updated":"2025-05-30T18:50:03Z","snapshot_observed_at":"2026-08-06T19:11:25.817469Z","submitted_at":"2025-04-12T11:44:47Z","title":"Parameterized Synthetic Text Generation with SimpleStories","version":3},"cited_work":{"arxiv_id":"2504.09184","doi":"10.48550/arxiv.2504.09184","metadata_source":"arxiv_reference","pith_arxiv_id":"2504.09184","snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"Parameterized Synthetic Text Generation with SimpleStories.arXiv preprint arXiv:2504.09184, 2025.https://arxiv.org/abs/2504.09184","venue":"ArXiv.org","work_id":"e8297923-01e4-4cbc-a7ed-0be3d60b4ed4","year":2025},"citing_paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse","version":1},"reference_index":60,"source":"arxiv_source","source_observed_at":"2026-05-12T03:58:11.179607Z"},"links":{"cited_paper":"/paper/2504.09184","citing_paper":"/paper/2605.09995"},"observation_digest":"sha256:381144f3ac00365467344033aa1d161807880cac465f4fa824c9c3f75d4028f1","observation_id":"bd5de481-0279-4cee-b58b-8fef13db4aeb","resolution":{"observed_at":"2026-05-12T06:46:51.857647Z","resolver_source":"arxiv_id","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-06T06:34:29.942622+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"state":"measured"}}],"paper":{"arxiv_id":"2605.09995","last_updated":"2026-05-11T05:11:04Z","latest_version":1,"primary_category":"cs.CL","snapshot_observed_at":"2026-07-06T23:21:59.096464Z","submitted_at":"2026-05-11T05:11:04Z","title":"Annotations Mitigate Post-Training Mode Collapse"},"reference_resolution":{"displayed":60,"state_counts":{"malformed_identifier":0,"metadata_mismatch":27,"parse_uncertain":0,"unresolved":1,"verified_exact":8,"verified_fuzzy":24},"total_outbound_references":60},"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":"2026-08-06T06:34:29.942622+00:00","source":"crossref"},{"observed_at":"2026-08-06T06:34:23.284952+00:00","source":"retraction_watch"}],"thesis":"As of 7 August 2026, this Paper Citation Record lists 60 of 60 outbound references and 0 inbound Pith citation observations for arXiv:2605.09995."}