{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TJ6Q2DYUBNTHYWKFIQJ53HCUWR","short_pith_number":"pith:TJ6Q2DYU","schema_version":"1.0","canonical_sha256":"9a7d0d0f140b667c59454413dd9c54b46e7ef2cfe24ab3f189b9d42ce6eba737","source":{"kind":"arxiv","id":"2505.07609","version":1},"attestation_state":"computed","paper":{"title":"TACOS: Temporally-aligned Audio CaptiOnS for Language-Audio Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Florian Schmid, Gerhard Widmer, Paul Primus","submitted_at":"2025-05-12T14:30:39Z","abstract_excerpt":"Learning to associate audio with textual descriptions is valuable for a range of tasks, including pretraining, zero-shot classification, audio retrieval, audio captioning, and text-conditioned audio generation. Existing contrastive language-audio pretrained models are typically trained using global, clip-level descriptions, which provide only weak temporal supervision. We hypothesize that CLAP-like language-audio models - particularly, if they are expected to produce frame-level embeddings - can benefit from a stronger temporal supervision. To confirm our hypothesis, we curate a novel dataset "},"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":"2505.07609","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2025-05-12T14:30:39Z","cross_cats_sorted":["cs.LG","cs.SD"],"title_canon_sha256":"9e5e1fcb75171cfea7aafc6616f393645103b63e13383bf2415e2caf29345dad","abstract_canon_sha256":"7444bc1788410c950270fa14c9b87482fa005660c3613f506af33f2261b88c7e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:01:50.592265Z","signature_b64":"ix0fIFsU2KEgV06T0A7S41m5qYa8nkjTev5p0Emz77NGStFi39rEaW5unaphZeze/t1ORVGS4mcoi296oPtbBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a7d0d0f140b667c59454413dd9c54b46e7ef2cfe24ab3f189b9d42ce6eba737","last_reissued_at":"2026-07-05T11:01:50.591699Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:01:50.591699Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TACOS: Temporally-aligned Audio CaptiOnS for Language-Audio Pretraining","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Florian Schmid, Gerhard Widmer, Paul Primus","submitted_at":"2025-05-12T14:30:39Z","abstract_excerpt":"Learning to associate audio with textual descriptions is valuable for a range of tasks, including pretraining, zero-shot classification, audio retrieval, audio captioning, and text-conditioned audio generation. Existing contrastive language-audio pretrained models are typically trained using global, clip-level descriptions, which provide only weak temporal supervision. We hypothesize that CLAP-like language-audio models - particularly, if they are expected to produce frame-level embeddings - can benefit from a stronger temporal supervision. To confirm our hypothesis, we curate a novel dataset "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.07609","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/2505.07609/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":"2505.07609","created_at":"2026-07-05T11:01:50.591765+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.07609v1","created_at":"2026-07-05T11:01:50.591765+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.07609","created_at":"2026-07-05T11:01:50.591765+00:00"},{"alias_kind":"pith_short_12","alias_value":"TJ6Q2DYUBNTH","created_at":"2026-07-05T11:01:50.591765+00:00"},{"alias_kind":"pith_short_16","alias_value":"TJ6Q2DYUBNTHYWKF","created_at":"2026-07-05T11:01:50.591765+00:00"},{"alias_kind":"pith_short_8","alias_value":"TJ6Q2DYU","created_at":"2026-07-05T11:01:50.591765+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.17417","citing_title":"A Closer Look at Failure Modes in Temporal Understanding of Large Audio-Language Models","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27741","citing_title":"Escape the Language Prior: Mitigating Late-Stage Modality Collapse in Audio Reasoning via Modality-Aware Policy Optimization","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR","json":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR.json","graph_json":"https://pith.science/api/pith-number/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/graph.json","events_json":"https://pith.science/api/pith-number/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/events.json","paper":"https://pith.science/paper/TJ6Q2DYU"},"agent_actions":{"view_html":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR","download_json":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR.json","view_paper":"https://pith.science/paper/TJ6Q2DYU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.07609&json=true","fetch_graph":"https://pith.science/api/pith-number/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/graph.json","fetch_events":"https://pith.science/api/pith-number/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/action/storage_attestation","attest_author":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/action/author_attestation","sign_citation":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/action/citation_signature","submit_replication":"https://pith.science/pith/TJ6Q2DYUBNTHYWKFIQJ53HCUWR/action/replication_record"}},"created_at":"2026-07-05T11:01:50.591765+00:00","updated_at":"2026-07-05T11:01:50.591765+00:00"}