{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:DJKLGBTKIYJI2TYUIXSOSGJZTY","short_pith_number":"pith:DJKLGBTK","schema_version":"1.0","canonical_sha256":"1a54b3066a46128d4f1445e4e919399e01a7ad841f61fdd6a976418ebf9c64dd","source":{"kind":"arxiv","id":"2607.26410","version":1},"attestation_state":"computed","paper":{"title":"Voice Memory for Agentic Speech Recognition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Boris Ginsburg, Chao-Han Huck Yang, Jagadeesh Balam, Piotr Zelasko, Zhehuai Chen, Zih-Ching Chen","submitted_at":"2026-07-29T02:42:56Z","abstract_excerpt":"We present Voice Memory, a inference-only scheme for agentic speech recognition: at stream time, a frozen corrector reads a single per-domain memory.md and decides per utterance whether to act on the hypothesis or abstain and keep the 1-best. Asynchronously, a score-gated optimizer revises that file through bounded edits, accepting an edit only when it strictly improves a held-out score. Extended from classical ASR-LM framework, we refer this split the listener-thinker architecture; the two roles are coupled only through the memory, so no weights change and the learned skill stays auditable an"},"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":"2607.26410","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-07-29T02:42:56Z","cross_cats_sorted":["cs.AI","cs.SD","eess.AS"],"title_canon_sha256":"2896278fedfafba1ea3c6d14ebd1776671e505728d6a8ca97347d5efc44bc3a4","abstract_canon_sha256":"53247b6320ff79a6c9656151d3de6fa14bdf2aeccf23e6894e40443d33ac0e8f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1a54b3066a46128d4f1445e4e919399e01a7ad841f61fdd6a976418ebf9c64dd","last_reissued_at":"2026-07-30T01:18:20.924742Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-30T01:18:20.924742Z"},"graph_snapshot":{"paper":{"title":"Voice Memory for Agentic Speech Recognition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Boris Ginsburg, Chao-Han Huck Yang, Jagadeesh Balam, Piotr Zelasko, Zhehuai Chen, Zih-Ching Chen","submitted_at":"2026-07-29T02:42:56Z","abstract_excerpt":"We present Voice Memory, a inference-only scheme for agentic speech recognition: at stream time, a frozen corrector reads a single per-domain memory.md and decides per utterance whether to act on the hypothesis or abstain and keep the 1-best. Asynchronously, a score-gated optimizer revises that file through bounded edits, accepting an edit only when it strictly improves a held-out score. Extended from classical ASR-LM framework, we refer this split the listener-thinker architecture; the two roles are coupled only through the memory, so no weights change and the learned skill stays auditable an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.26410","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/2607.26410/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":"2607.26410","created_at":"2026-07-30T01:18:20.929980+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.26410v1","created_at":"2026-07-30T01:18:20.929980+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.26410","created_at":"2026-07-30T01:18:20.929980+00:00"},{"alias_kind":"pith_short_12","alias_value":"DJKLGBTKIYJI","created_at":"2026-07-30T01:18:20.929980+00:00"},{"alias_kind":"pith_short_16","alias_value":"DJKLGBTKIYJI2TYU","created_at":"2026-07-30T01:18:20.929980+00:00"},{"alias_kind":"pith_short_8","alias_value":"DJKLGBTK","created_at":"2026-07-30T01:18:20.929980+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/DJKLGBTKIYJI2TYUIXSOSGJZTY","json":"https://pith.science/pith/DJKLGBTKIYJI2TYUIXSOSGJZTY.json","graph_json":"https://pith.science/api/pith-number/DJKLGBTKIYJI2TYUIXSOSGJZTY/graph.json","events_json":"https://pith.science/api/pith-number/DJKLGBTKIYJI2TYUIXSOSGJZTY/events.json","paper":"https://pith.science/paper/DJKLGBTK"},"agent_actions":{"view_html":"https://pith.science/pith/DJKLGBTKIYJI2TYUIXSOSGJZTY","download_json":"https://pith.science/pith/DJKLGBTKIYJI2TYUIXSOSGJZTY.json","view_paper":"https://pith.science/paper/DJKLGBTK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.26410&json=true","fetch_graph":"https://pith.science/api/pith-number/DJKLGBTKIYJI2TYUIXSOSGJZTY/graph.json","fetch_events":"https://pith.science/api/pith-number/DJKLGBTKIYJI2TYUIXSOSGJZTY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DJKLGBTKIYJI2TYUIXSOSGJZTY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DJKLGBTKIYJI2TYUIXSOSGJZTY/action/storage_attestation","attest_author":"https://pith.science/pith/DJKLGBTKIYJI2TYUIXSOSGJZTY/action/author_attestation","sign_citation":"https://pith.science/pith/DJKLGBTKIYJI2TYUIXSOSGJZTY/action/citation_signature","submit_replication":"https://pith.science/pith/DJKLGBTKIYJI2TYUIXSOSGJZTY/action/replication_record"}},"created_at":"2026-07-30T01:18:20.929980+00:00","updated_at":"2026-07-30T01:18:20.929980+00:00"}