{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HCO45N7QQ47OU5CYUVXT7UW3XF","short_pith_number":"pith:HCO45N7Q","schema_version":"1.0","canonical_sha256":"389dceb7f0873eea7458a56f3fd2dbb9655562b94c2f8ca109ee0ba53e701e8a","source":{"kind":"arxiv","id":"2404.00685","version":2},"attestation_state":"computed","paper":{"title":"Scaling Properties of Speech Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.NE"],"primary_cat":"eess.AS","authors_text":"Ricard Marxer, Santiago Cuervo","submitted_at":"2024-03-31T13:30:12Z","abstract_excerpt":"Speech Language Models (SLMs) aim to learn language from raw audio, without textual resources. Despite significant advances, our current models exhibit weak syntax and semantic abilities. However, if the scaling properties of neural language models hold for the speech modality, these abilities will improve as the amount of compute used for training increases. In this paper, we use models of this scaling behavior to estimate the scale at which our current methods will yield a SLM with the English proficiency of text-based Large Language Models (LLMs). We establish a strong correlation between p"},"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":"2404.00685","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2024-03-31T13:30:12Z","cross_cats_sorted":["cs.AI","cs.CL","cs.NE"],"title_canon_sha256":"3b20eeb5ef43ca747d2904ef55b12a0fe39866bd299a502726052935212c566a","abstract_canon_sha256":"bca0371413992aae256976cdb3915b057e997a036e3ec4ad53a64d011eabc040"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:48:12.918522Z","signature_b64":"SX/1O7l6cG4srARzl9AbmiyyM6jBSg34txMn27TNkQyKE17qFT5aRg+BjoLsR5SlFnN5yaXUTVOEkb++Ek+nBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"389dceb7f0873eea7458a56f3fd2dbb9655562b94c2f8ca109ee0ba53e701e8a","last_reissued_at":"2026-07-05T09:48:12.918099Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:48:12.918099Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scaling Properties of Speech Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.NE"],"primary_cat":"eess.AS","authors_text":"Ricard Marxer, Santiago Cuervo","submitted_at":"2024-03-31T13:30:12Z","abstract_excerpt":"Speech Language Models (SLMs) aim to learn language from raw audio, without textual resources. Despite significant advances, our current models exhibit weak syntax and semantic abilities. However, if the scaling properties of neural language models hold for the speech modality, these abilities will improve as the amount of compute used for training increases. In this paper, we use models of this scaling behavior to estimate the scale at which our current methods will yield a SLM with the English proficiency of text-based Large Language Models (LLMs). We establish a strong correlation between p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.00685","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/2404.00685/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":"2404.00685","created_at":"2026-07-05T09:48:12.918155+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.00685v2","created_at":"2026-07-05T09:48:12.918155+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.00685","created_at":"2026-07-05T09:48:12.918155+00:00"},{"alias_kind":"pith_short_12","alias_value":"HCO45N7QQ47O","created_at":"2026-07-05T09:48:12.918155+00:00"},{"alias_kind":"pith_short_16","alias_value":"HCO45N7QQ47OU5CY","created_at":"2026-07-05T09:48:12.918155+00:00"},{"alias_kind":"pith_short_8","alias_value":"HCO45N7Q","created_at":"2026-07-05T09:48:12.918155+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.04870","citing_title":"Aligning Multimodal Representations through an Information Bottleneck","ref_index":14,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF","json":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF.json","graph_json":"https://pith.science/api/pith-number/HCO45N7QQ47OU5CYUVXT7UW3XF/graph.json","events_json":"https://pith.science/api/pith-number/HCO45N7QQ47OU5CYUVXT7UW3XF/events.json","paper":"https://pith.science/paper/HCO45N7Q"},"agent_actions":{"view_html":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF","download_json":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF.json","view_paper":"https://pith.science/paper/HCO45N7Q","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.00685&json=true","fetch_graph":"https://pith.science/api/pith-number/HCO45N7QQ47OU5CYUVXT7UW3XF/graph.json","fetch_events":"https://pith.science/api/pith-number/HCO45N7QQ47OU5CYUVXT7UW3XF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF/action/storage_attestation","attest_author":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF/action/author_attestation","sign_citation":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF/action/citation_signature","submit_replication":"https://pith.science/pith/HCO45N7QQ47OU5CYUVXT7UW3XF/action/replication_record"}},"created_at":"2026-07-05T09:48:12.918155+00:00","updated_at":"2026-07-05T09:48:12.918155+00:00"}