{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:NNI66T4GFHKVTC3YU54A4O5CSH","short_pith_number":"pith:NNI66T4G","schema_version":"1.0","canonical_sha256":"6b51ef4f8629d5598b78a7780e3ba291ce574e3078ff2ef4d331422a55b2e415","source":{"kind":"arxiv","id":"2608.01560","version":1},"attestation_state":"computed","paper":{"title":"Discriminative Axis, Not Data Volume: What a Contrastive Corpus Teaches an Audio Embedding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"cs.CL","authors_text":"Abdul Basit Tonmoy","submitted_at":"2026-08-03T00:34:45Z","abstract_excerpt":"Scaling the corpus is the default remedy when a contrastive representation lacks an attribute. We report a case where it does nothing, and identify what does: adding a lexical-speech round to a frozen-base multimodal embedding model raises zero-shot keyword spotting by 76 points while reducing speech-emotion recognition by 14. The loss is not a capacity limit: fine-tuning on 7,442 clips from a prosody-controlled corpus recovers emotion past its pre-speech level at a five-point keyword cost. Nor is it data volume: 29,428 mined clips whose captions explicitly name emotions, at matched exposure, "},"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":"2608.01560","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-03T00:34:45Z","cross_cats_sorted":["cs.SD"],"title_canon_sha256":"90f00799018e2ac98503a4b7e2a8994e46aa5d1fadf8cd0416f6fcbc3720fe24","abstract_canon_sha256":"82d6ec675c79540ba2cd6b3324f35a036f28014d39f8f449b71debdb6b33cddd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-04T02:05:59.567702Z","signature_b64":"4ZFShmrrqArAMd/ejDFZPRvP1+sc1MYGcD0xM4IsL1SNSlQXU6sZvfpJS+9R+QO+j/2Kh1fxLKVyMgQNC17oCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6b51ef4f8629d5598b78a7780e3ba291ce574e3078ff2ef4d331422a55b2e415","last_reissued_at":"2026-08-04T02:05:59.566130Z","signature_status":"signed_v1","first_computed_at":"2026-08-04T02:05:59.566130Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Discriminative Axis, Not Data Volume: What a Contrastive Corpus Teaches an Audio Embedding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SD"],"primary_cat":"cs.CL","authors_text":"Abdul Basit Tonmoy","submitted_at":"2026-08-03T00:34:45Z","abstract_excerpt":"Scaling the corpus is the default remedy when a contrastive representation lacks an attribute. We report a case where it does nothing, and identify what does: adding a lexical-speech round to a frozen-base multimodal embedding model raises zero-shot keyword spotting by 76 points while reducing speech-emotion recognition by 14. The loss is not a capacity limit: fine-tuning on 7,442 clips from a prosody-controlled corpus recovers emotion past its pre-speech level at a five-point keyword cost. Nor is it data volume: 29,428 mined clips whose captions explicitly name emotions, at matched exposure, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.01560","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/2608.01560/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":"2608.01560","created_at":"2026-08-04T02:05:59.567553+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.01560v1","created_at":"2026-08-04T02:05:59.567553+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.01560","created_at":"2026-08-04T02:05:59.567553+00:00"},{"alias_kind":"pith_short_12","alias_value":"NNI66T4GFHKV","created_at":"2026-08-04T02:05:59.567553+00:00"},{"alias_kind":"pith_short_16","alias_value":"NNI66T4GFHKVTC3Y","created_at":"2026-08-04T02:05:59.567553+00:00"},{"alias_kind":"pith_short_8","alias_value":"NNI66T4G","created_at":"2026-08-04T02:05:59.567553+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/NNI66T4GFHKVTC3YU54A4O5CSH","json":"https://pith.science/pith/NNI66T4GFHKVTC3YU54A4O5CSH.json","graph_json":"https://pith.science/api/pith-number/NNI66T4GFHKVTC3YU54A4O5CSH/graph.json","events_json":"https://pith.science/api/pith-number/NNI66T4GFHKVTC3YU54A4O5CSH/events.json","paper":"https://pith.science/paper/NNI66T4G"},"agent_actions":{"view_html":"https://pith.science/pith/NNI66T4GFHKVTC3YU54A4O5CSH","download_json":"https://pith.science/pith/NNI66T4GFHKVTC3YU54A4O5CSH.json","view_paper":"https://pith.science/paper/NNI66T4G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.01560&json=true","fetch_graph":"https://pith.science/api/pith-number/NNI66T4GFHKVTC3YU54A4O5CSH/graph.json","fetch_events":"https://pith.science/api/pith-number/NNI66T4GFHKVTC3YU54A4O5CSH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NNI66T4GFHKVTC3YU54A4O5CSH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NNI66T4GFHKVTC3YU54A4O5CSH/action/storage_attestation","attest_author":"https://pith.science/pith/NNI66T4GFHKVTC3YU54A4O5CSH/action/author_attestation","sign_citation":"https://pith.science/pith/NNI66T4GFHKVTC3YU54A4O5CSH/action/citation_signature","submit_replication":"https://pith.science/pith/NNI66T4GFHKVTC3YU54A4O5CSH/action/replication_record"}},"created_at":"2026-08-04T02:05:59.567553+00:00","updated_at":"2026-08-04T02:05:59.567553+00:00"}