{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BLWQ6IXIW4GOHHR2BYGGYAAPOU","short_pith_number":"pith:BLWQ6IXI","schema_version":"1.0","canonical_sha256":"0aed0f22e8b70ce39e3a0e0c6c000f75359df38f6eac1cfc4529e060795e3168","source":{"kind":"arxiv","id":"2506.06096","version":1},"attestation_state":"computed","paper":{"title":"Label-Context-Dependent Internal Language Model Estimation for CTC","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG","eess.AS"],"primary_cat":"cs.SD","authors_text":"Hermann Ney, Minh-Nghia Phan, Ralf Schl\\\"uter, Zijian Yang","submitted_at":"2025-06-06T13:54:43Z","abstract_excerpt":"Although connectionist temporal classification (CTC) has the label context independence assumption, it can still implicitly learn a context-dependent internal language model (ILM) due to modern powerful encoders. In this work, we investigate the implicit context dependency modeled in the ILM of CTC. To this end, we propose novel context-dependent ILM estimation methods for CTC based on knowledge distillation (KD) with theoretical justifications. Furthermore, we introduce two regularization methods for KD. We conduct experiments on Librispeech and TED-LIUM Release 2 datasets for in-domain and c"},"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":"2506.06096","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.SD","submitted_at":"2025-06-06T13:54:43Z","cross_cats_sorted":["cs.CL","cs.LG","eess.AS"],"title_canon_sha256":"1b012e9ac8ada216615030b3475edca471b44433d2ec3ff8743795417e322a2a","abstract_canon_sha256":"09fa954b2d2ff39b543d6b70ba59571b147621acfd0287be75e1753416132972"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:17:25.531808Z","signature_b64":"revDgzVQODjqcNJ64UMUkBzkLvze6u/55FHKmlQlVMJRChv5ZJ+aJU5T8DwYE/bwKgGBf0Eg79Vwh0fvjhU2DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0aed0f22e8b70ce39e3a0e0c6c000f75359df38f6eac1cfc4529e060795e3168","last_reissued_at":"2026-07-05T11:17:25.531240Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:17:25.531240Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Label-Context-Dependent Internal Language Model Estimation for CTC","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG","eess.AS"],"primary_cat":"cs.SD","authors_text":"Hermann Ney, Minh-Nghia Phan, Ralf Schl\\\"uter, Zijian Yang","submitted_at":"2025-06-06T13:54:43Z","abstract_excerpt":"Although connectionist temporal classification (CTC) has the label context independence assumption, it can still implicitly learn a context-dependent internal language model (ILM) due to modern powerful encoders. In this work, we investigate the implicit context dependency modeled in the ILM of CTC. To this end, we propose novel context-dependent ILM estimation methods for CTC based on knowledge distillation (KD) with theoretical justifications. Furthermore, we introduce two regularization methods for KD. We conduct experiments on Librispeech and TED-LIUM Release 2 datasets for in-domain and c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.06096","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/2506.06096/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":"2506.06096","created_at":"2026-07-05T11:17:25.531306+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.06096v1","created_at":"2026-07-05T11:17:25.531306+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.06096","created_at":"2026-07-05T11:17:25.531306+00:00"},{"alias_kind":"pith_short_12","alias_value":"BLWQ6IXIW4GO","created_at":"2026-07-05T11:17:25.531306+00:00"},{"alias_kind":"pith_short_16","alias_value":"BLWQ6IXIW4GOHHR2","created_at":"2026-07-05T11:17:25.531306+00:00"},{"alias_kind":"pith_short_8","alias_value":"BLWQ6IXI","created_at":"2026-07-05T11:17:25.531306+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.06096","citing_title":"Label-Context-Dependent Internal Language Model Estimation for CTC","ref_index":2,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU","json":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU.json","graph_json":"https://pith.science/api/pith-number/BLWQ6IXIW4GOHHR2BYGGYAAPOU/graph.json","events_json":"https://pith.science/api/pith-number/BLWQ6IXIW4GOHHR2BYGGYAAPOU/events.json","paper":"https://pith.science/paper/BLWQ6IXI"},"agent_actions":{"view_html":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU","download_json":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU.json","view_paper":"https://pith.science/paper/BLWQ6IXI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.06096&json=true","fetch_graph":"https://pith.science/api/pith-number/BLWQ6IXIW4GOHHR2BYGGYAAPOU/graph.json","fetch_events":"https://pith.science/api/pith-number/BLWQ6IXIW4GOHHR2BYGGYAAPOU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU/action/storage_attestation","attest_author":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU/action/author_attestation","sign_citation":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU/action/citation_signature","submit_replication":"https://pith.science/pith/BLWQ6IXIW4GOHHR2BYGGYAAPOU/action/replication_record"}},"created_at":"2026-07-05T11:17:25.531306+00:00","updated_at":"2026-07-05T11:17:25.531306+00:00"}