{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:JOLJIJRDWBLQ6IQCEPN3HEE2JO","short_pith_number":"pith:JOLJIJRD","schema_version":"1.0","canonical_sha256":"4b96942623b0570f220223dbb3909a4b89357a72b58e5e118dd02864ab53b9e5","source":{"kind":"arxiv","id":"2311.16822","version":2},"attestation_state":"computed","paper":{"title":"Large Language Models Suffer From Their Own Output: An Analysis of the Self-Consuming Training Loop","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.NE"],"primary_cat":"cs.LG","authors_text":"Dominik Sobania, Franz Rothlauf, Martin Briesch","submitted_at":"2023-11-28T14:36:43Z","abstract_excerpt":"Large Language Models (LLM) are already widely used to generate content for a variety of online platforms. As we are not able to safely distinguish LLM-generated content from human-produced content, LLM-generated content is used to train the next generation of LLMs, giving rise to a self-consuming training loop. From the image generation domain we know that such a self-consuming training loop reduces both quality and diversity of images finally ending in a model collapse. However, it is unclear whether this alarming effect can also be observed for LLMs. Therefore, we present the first study in"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2311.16822","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-28T14:36:43Z","cross_cats_sorted":["cs.CL","cs.NE"],"title_canon_sha256":"d2b6e5fab5bbccfed4991713a614e656f01ad61fc039dbcf7672b658af0eb61d","abstract_canon_sha256":"503df6b42e59687066adb9e880ff2c7af138180e95592f1454d435b0bd5d8703"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:32:33.328550Z","signature_b64":"beQOXgHoTnIKMs0D/lCvzsW9m7f7TTfu7tdq4s7zw7/tt8gzdvKog8M9KN+CpnciRVzWLspSzneQe0mKsc/vDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4b96942623b0570f220223dbb3909a4b89357a72b58e5e118dd02864ab53b9e5","last_reissued_at":"2026-07-05T08:32:33.327948Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:32:33.327948Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Large Language Models Suffer From Their Own Output: An Analysis of the Self-Consuming Training Loop","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.NE"],"primary_cat":"cs.LG","authors_text":"Dominik Sobania, Franz Rothlauf, Martin Briesch","submitted_at":"2023-11-28T14:36:43Z","abstract_excerpt":"Large Language Models (LLM) are already widely used to generate content for a variety of online platforms. As we are not able to safely distinguish LLM-generated content from human-produced content, LLM-generated content is used to train the next generation of LLMs, giving rise to a self-consuming training loop. From the image generation domain we know that such a self-consuming training loop reduces both quality and diversity of images finally ending in a model collapse. However, it is unclear whether this alarming effect can also be observed for LLMs. Therefore, we present the first study in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.16822","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2311.16822/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2311.16822","created_at":"2026-07-05T08:32:33.328008+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.16822v2","created_at":"2026-07-05T08:32:33.328008+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.16822","created_at":"2026-07-05T08:32:33.328008+00:00"},{"alias_kind":"pith_short_12","alias_value":"JOLJIJRDWBLQ","created_at":"2026-07-05T08:32:33.328008+00:00"},{"alias_kind":"pith_short_16","alias_value":"JOLJIJRDWBLQ6IQC","created_at":"2026-07-05T08:32:33.328008+00:00"},{"alias_kind":"pith_short_8","alias_value":"JOLJIJRD","created_at":"2026-07-05T08:32:33.328008+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.00334","citing_title":"Isolating LLM Lexical Bias: A Curation-Free Triangulated Metric for Preference-Stage Learning","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28438","citing_title":"When AI Reviews Its Own Code: Recursive Self-Training Collapse in Code LLMs","ref_index":150,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23054","citing_title":"Model Collapse as Cultural Evolution","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20602","citing_title":"Self-Training Doesn't Flatten Language -- It Restructures It: Surface Markers Amplify While Deep Syntax Dies","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20279","citing_title":"The Economics of Model Collapse: Equilibrium, Welfare, and Optimal Provenance Subsidies in Synthetic Data Markets","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20151","citing_title":"When Does Model Collapse Occur in Structured Interactive Learning?","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2511.09416","citing_title":"Transformer Semantic Genetic Programming for d-dimensional Symbolic Regression Problems","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO","json":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO.json","graph_json":"https://pith.science/api/pith-number/JOLJIJRDWBLQ6IQCEPN3HEE2JO/graph.json","events_json":"https://pith.science/api/pith-number/JOLJIJRDWBLQ6IQCEPN3HEE2JO/events.json","paper":"https://pith.science/paper/JOLJIJRD"},"agent_actions":{"view_html":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO","download_json":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO.json","view_paper":"https://pith.science/paper/JOLJIJRD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.16822&json=true","fetch_graph":"https://pith.science/api/pith-number/JOLJIJRDWBLQ6IQCEPN3HEE2JO/graph.json","fetch_events":"https://pith.science/api/pith-number/JOLJIJRDWBLQ6IQCEPN3HEE2JO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO/action/storage_attestation","attest_author":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO/action/author_attestation","sign_citation":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO/action/citation_signature","submit_replication":"https://pith.science/pith/JOLJIJRDWBLQ6IQCEPN3HEE2JO/action/replication_record"}},"created_at":"2026-07-05T08:32:33.328008+00:00","updated_at":"2026-07-05T08:32:33.328008+00:00"}