{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:N2BYTL3CD3SWV3CU2ASNFGXMQ2","short_pith_number":"pith:N2BYTL3C","schema_version":"1.0","canonical_sha256":"6e8389af621ee56aec54d024d29aec86b4745a485a9324d648ebcbb5a0c41643","source":{"kind":"arxiv","id":"2010.12626","version":1},"attestation_state":"computed","paper":{"title":"Topic Modeling with Contextualized Word Representation Clusters","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"David Mimno, Laure Thompson","submitted_at":"2020-10-23T19:16:59Z","abstract_excerpt":"Clustering token-level contextualized word representations produces output that shares many similarities with topic models for English text collections. Unlike clusterings of vocabulary-level word embeddings, the resulting models more naturally capture polysemy and can be used as a way of organizing documents. We evaluate token clusterings trained from several different output layers of popular contextualized language models. We find that BERT and GPT-2 produce high quality clusterings, but RoBERTa does not. These cluster models are simple, reliable, and can perform as well as, if not better t"},"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":"2010.12626","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2020-10-23T19:16:59Z","cross_cats_sorted":[],"title_canon_sha256":"a5ba4cf2595dd9304ed0a3eaac7d431cb8c9576b0b3c59d69b6ec8953a73aff4","abstract_canon_sha256":"b5b0b5748c89ae7f3bba9d8d8a47eea2acd86b520f19ab4f68113ac3b5cd78ca"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:45:46.578521Z","signature_b64":"7cPP7Ke0v6EZ8RjGbZlqdhwRx64Hh0e04JOqgm9qrA1nSku2UibcS9/uEgr3c/iN/azBH9SNg0G/c9d0XYbVCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6e8389af621ee56aec54d024d29aec86b4745a485a9324d648ebcbb5a0c41643","last_reissued_at":"2026-07-05T01:45:46.578117Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:45:46.578117Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Topic Modeling with Contextualized Word Representation Clusters","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"David Mimno, Laure Thompson","submitted_at":"2020-10-23T19:16:59Z","abstract_excerpt":"Clustering token-level contextualized word representations produces output that shares many similarities with topic models for English text collections. Unlike clusterings of vocabulary-level word embeddings, the resulting models more naturally capture polysemy and can be used as a way of organizing documents. We evaluate token clusterings trained from several different output layers of popular contextualized language models. We find that BERT and GPT-2 produce high quality clusterings, but RoBERTa does not. These cluster models are simple, reliable, and can perform as well as, if not better t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.12626","kind":"arxiv","version":1},"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/2010.12626/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":"2010.12626","created_at":"2026-07-05T01:45:46.578178+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.12626v1","created_at":"2026-07-05T01:45:46.578178+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.12626","created_at":"2026-07-05T01:45:46.578178+00:00"},{"alias_kind":"pith_short_12","alias_value":"N2BYTL3CD3SW","created_at":"2026-07-05T01:45:46.578178+00:00"},{"alias_kind":"pith_short_16","alias_value":"N2BYTL3CD3SWV3CU","created_at":"2026-07-05T01:45:46.578178+00:00"},{"alias_kind":"pith_short_8","alias_value":"N2BYTL3C","created_at":"2026-07-05T01:45:46.578178+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2","json":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2.json","graph_json":"https://pith.science/api/pith-number/N2BYTL3CD3SWV3CU2ASNFGXMQ2/graph.json","events_json":"https://pith.science/api/pith-number/N2BYTL3CD3SWV3CU2ASNFGXMQ2/events.json","paper":"https://pith.science/paper/N2BYTL3C"},"agent_actions":{"view_html":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2","download_json":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2.json","view_paper":"https://pith.science/paper/N2BYTL3C","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.12626&json=true","fetch_graph":"https://pith.science/api/pith-number/N2BYTL3CD3SWV3CU2ASNFGXMQ2/graph.json","fetch_events":"https://pith.science/api/pith-number/N2BYTL3CD3SWV3CU2ASNFGXMQ2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2/action/storage_attestation","attest_author":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2/action/author_attestation","sign_citation":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2/action/citation_signature","submit_replication":"https://pith.science/pith/N2BYTL3CD3SWV3CU2ASNFGXMQ2/action/replication_record"}},"created_at":"2026-07-05T01:45:46.578178+00:00","updated_at":"2026-07-05T01:45:46.578178+00:00"}