{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:QCZEL3XUS7Q7VHJOSW5M5OIR34","short_pith_number":"pith:QCZEL3XU","schema_version":"1.0","canonical_sha256":"80b245eef497e1fa9d2e95baceb911df36d3ba8af2feb8ef794f36e7d86dff6b","source":{"kind":"arxiv","id":"2606.26824","version":1},"attestation_state":"computed","paper":{"title":"wav2tok 2.0: Scalable Audio Tokenization Maintaining Explicit Pairwise Token Alignment for Efficient Audio Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["eess.AS"],"primary_cat":"cs.SD","authors_text":"Adhiraj Banerjee, Vipul Arora","submitted_at":"2026-06-25T10:04:35Z","abstract_excerpt":"Learning discrete speech representations that preserve similarity across variable-length utterances is central to query-by-example spoken term detection (QbE-STD). While wav2tok introduced CTC-based sequence alignment to enforce token consistency, its tightly coupled clustering and alignment training recipe limits scalability. We propose wav2tok 2.0, a scalable alignment-aware speech tokenizer built on the BEST-STD backbone. wav2tok 2.0 employs staged training, first learning discriminative, speaker-invariant representations via contrastive learning and vector quantization, and then enforcing "},"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":"2606.26824","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SD","submitted_at":"2026-06-25T10:04:35Z","cross_cats_sorted":["eess.AS"],"title_canon_sha256":"4c51152d792fd253303a01119088bac20ad58f4fcf598667f56f38edaf1b3a6b","abstract_canon_sha256":"5478fe25a656c33c634f833ee46832150179eeaa19a06af7aa754bb5ab5440db"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-26T01:16:00.900697Z","signature_b64":"Cr+maQnaQ/apngmZ7N9yTIN0GScw/vx9pzpGT1ICkKEKuRn5k3+GT5+irBd0GCFeaK9r9pmWGE4AQAjGzn3PAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"80b245eef497e1fa9d2e95baceb911df36d3ba8af2feb8ef794f36e7d86dff6b","last_reissued_at":"2026-06-26T01:16:00.900302Z","signature_status":"signed_v1","first_computed_at":"2026-06-26T01:16:00.900302Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"wav2tok 2.0: Scalable Audio Tokenization Maintaining Explicit Pairwise Token Alignment for Efficient Audio Retrieval","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["eess.AS"],"primary_cat":"cs.SD","authors_text":"Adhiraj Banerjee, Vipul Arora","submitted_at":"2026-06-25T10:04:35Z","abstract_excerpt":"Learning discrete speech representations that preserve similarity across variable-length utterances is central to query-by-example spoken term detection (QbE-STD). While wav2tok introduced CTC-based sequence alignment to enforce token consistency, its tightly coupled clustering and alignment training recipe limits scalability. We propose wav2tok 2.0, a scalable alignment-aware speech tokenizer built on the BEST-STD backbone. wav2tok 2.0 employs staged training, first learning discriminative, speaker-invariant representations via contrastive learning and vector quantization, and then enforcing "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.26824","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/2606.26824/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":"2606.26824","created_at":"2026-06-26T01:16:00.900367+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.26824v1","created_at":"2026-06-26T01:16:00.900367+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.26824","created_at":"2026-06-26T01:16:00.900367+00:00"},{"alias_kind":"pith_short_12","alias_value":"QCZEL3XUS7Q7","created_at":"2026-06-26T01:16:00.900367+00:00"},{"alias_kind":"pith_short_16","alias_value":"QCZEL3XUS7Q7VHJO","created_at":"2026-06-26T01:16:00.900367+00:00"},{"alias_kind":"pith_short_8","alias_value":"QCZEL3XU","created_at":"2026-06-26T01:16:00.900367+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2606.26824","citing_title":"wav2tok 2.0: Scalable Audio Tokenization Maintaining Explicit Pairwise Token Alignment for Efficient Audio Retrieval","ref_index":2,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34","json":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34.json","graph_json":"https://pith.science/api/pith-number/QCZEL3XUS7Q7VHJOSW5M5OIR34/graph.json","events_json":"https://pith.science/api/pith-number/QCZEL3XUS7Q7VHJOSW5M5OIR34/events.json","paper":"https://pith.science/paper/QCZEL3XU"},"agent_actions":{"view_html":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34","download_json":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34.json","view_paper":"https://pith.science/paper/QCZEL3XU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.26824&json=true","fetch_graph":"https://pith.science/api/pith-number/QCZEL3XUS7Q7VHJOSW5M5OIR34/graph.json","fetch_events":"https://pith.science/api/pith-number/QCZEL3XUS7Q7VHJOSW5M5OIR34/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34/action/storage_attestation","attest_author":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34/action/author_attestation","sign_citation":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34/action/citation_signature","submit_replication":"https://pith.science/pith/QCZEL3XUS7Q7VHJOSW5M5OIR34/action/replication_record"}},"created_at":"2026-06-26T01:16:00.900367+00:00","updated_at":"2026-06-26T01:16:00.900367+00:00"}