{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:G26LJGP2IT6L733M6IZRML6XRH","short_pith_number":"pith:G26LJGP2","schema_version":"1.0","canonical_sha256":"36bcb499fa44fcbfef6cf233162fd789eec07a1c2cc21e7bd7a1fb0ebac326b4","source":{"kind":"arxiv","id":"2602.17907","version":2},"attestation_state":"computed","paper":{"title":"DSL-Topic: Improving Topic Modeling by Distilling Soft Labelsfrom Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Amirhossein Abaskohi, Chuyuan Li, Gabriel Murray, Giuseppe Carenini, Raymond Li","submitted_at":"2026-02-20T00:12:04Z","abstract_excerpt":"Traditional neural topic models are typically optimized by reconstructing the document's Bag-of-Words (BoW) representations, overlooking contextual information and struggling with data sparsity. In this work, we introduce a novel topic model training framework by Distilling Soft Labels (DSL) from Language Models (LMs). To construct the contextually enriched reconstruction signals, we project the next token probabilities, conditioned on a specialized prompt, onto a pre-defined vocabulary, and train the topic models to reconstruct the soft labels using the LM hidden states. This produces higher-"},"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":"2602.17907","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-20T00:12:04Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ff149898a957bb53dd585a0b4390f2cf731b419f71205138e181cc4733f4ff64","abstract_canon_sha256":"c7fb8820cd6139316dbac167ab8007113129d769fb3509d626bd5f7007f7f7f0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-04T01:08:45.802618Z","signature_b64":"6OgYrezsXgKx45LSFlS0S5srt5xR+uupJyURrrFoVkedsLICdqGXzfPgBIXvUc5ktG341G24zhuO9+MgVaq2Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"36bcb499fa44fcbfef6cf233162fd789eec07a1c2cc21e7bd7a1fb0ebac326b4","last_reissued_at":"2026-06-04T01:08:45.801787Z","signature_status":"signed_v1","first_computed_at":"2026-06-04T01:08:45.801787Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DSL-Topic: Improving Topic Modeling by Distilling Soft Labelsfrom Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Amirhossein Abaskohi, Chuyuan Li, Gabriel Murray, Giuseppe Carenini, Raymond Li","submitted_at":"2026-02-20T00:12:04Z","abstract_excerpt":"Traditional neural topic models are typically optimized by reconstructing the document's Bag-of-Words (BoW) representations, overlooking contextual information and struggling with data sparsity. In this work, we introduce a novel topic model training framework by Distilling Soft Labels (DSL) from Language Models (LMs). To construct the contextually enriched reconstruction signals, we project the next token probabilities, conditioned on a specialized prompt, onto a pre-defined vocabulary, and train the topic models to reconstruct the soft labels using the LM hidden states. This produces higher-"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.17907","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/2602.17907/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":"2602.17907","created_at":"2026-06-04T01:08:45.801890+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.17907v2","created_at":"2026-06-04T01:08:45.801890+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.17907","created_at":"2026-06-04T01:08:45.801890+00:00"},{"alias_kind":"pith_short_12","alias_value":"G26LJGP2IT6L","created_at":"2026-06-04T01:08:45.801890+00:00"},{"alias_kind":"pith_short_16","alias_value":"G26LJGP2IT6L733M","created_at":"2026-06-04T01:08:45.801890+00:00"},{"alias_kind":"pith_short_8","alias_value":"G26LJGP2","created_at":"2026-06-04T01:08:45.801890+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/G26LJGP2IT6L733M6IZRML6XRH","json":"https://pith.science/pith/G26LJGP2IT6L733M6IZRML6XRH.json","graph_json":"https://pith.science/api/pith-number/G26LJGP2IT6L733M6IZRML6XRH/graph.json","events_json":"https://pith.science/api/pith-number/G26LJGP2IT6L733M6IZRML6XRH/events.json","paper":"https://pith.science/paper/G26LJGP2"},"agent_actions":{"view_html":"https://pith.science/pith/G26LJGP2IT6L733M6IZRML6XRH","download_json":"https://pith.science/pith/G26LJGP2IT6L733M6IZRML6XRH.json","view_paper":"https://pith.science/paper/G26LJGP2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.17907&json=true","fetch_graph":"https://pith.science/api/pith-number/G26LJGP2IT6L733M6IZRML6XRH/graph.json","fetch_events":"https://pith.science/api/pith-number/G26LJGP2IT6L733M6IZRML6XRH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G26LJGP2IT6L733M6IZRML6XRH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G26LJGP2IT6L733M6IZRML6XRH/action/storage_attestation","attest_author":"https://pith.science/pith/G26LJGP2IT6L733M6IZRML6XRH/action/author_attestation","sign_citation":"https://pith.science/pith/G26LJGP2IT6L733M6IZRML6XRH/action/citation_signature","submit_replication":"https://pith.science/pith/G26LJGP2IT6L733M6IZRML6XRH/action/replication_record"}},"created_at":"2026-06-04T01:08:45.801890+00:00","updated_at":"2026-06-04T01:08:45.801890+00:00"}