{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YHWPJFFREDHJ7FHOS7W4RAVTSE","short_pith_number":"pith:YHWPJFFR","schema_version":"1.0","canonical_sha256":"c1ecf494b120ce9f94ee97edc882b3913091786d5272d7a050865043fee8990b","source":{"kind":"arxiv","id":"2606.00460","version":1},"attestation_state":"computed","paper":{"title":"SALSA: Speech Aware LLM Adaptation via Learned Steering Activation Vectors","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["eess.AS"],"primary_cat":"cs.CL","authors_text":"Argyrios Gerogiannis, Chang D. Yoo, Haolong Zheng, Julia Hockenmaier, Mark A. Hasegawa-Johnson, Yekaterina Yegorova","submitted_at":"2026-05-30T00:54:53Z","abstract_excerpt":"Speech-aware large language models often generalize poorly to out-of-domain settings. We propose SALSA (Speech-Aware LLM Adaptation via Learned Steering Activations), a lightweight adaptation method that learns layer-wise steering vectors. Unlike commonly used steering approaches that rely on contrastive activation differences, SALSA directly optimizes steering vectors using a supervised objective. Across children's speech, multilingual speech, and Mandarin-English code-switching benchmarks, SALSA substantially improves performance over zero-shot inference and speech in-context learning baseli"},"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":"2606.00460","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-30T00:54:53Z","cross_cats_sorted":["eess.AS"],"title_canon_sha256":"ed5b002b66cca73e151eb9162506f605949076d201fb0b4471ef27b583d34852","abstract_canon_sha256":"57f3c3718bc5ba41dd246c1a36b51205fd1939bb152194f57464d705fe14c586"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:55.339647Z","signature_b64":"O9YC7xb4QC5boWB7u80k1aXO0ftx+/J0s/oLISatBwiNHdgYEFg87TsZb1O1f7qrr/gfewTaGCbwH+E0nuKlAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c1ecf494b120ce9f94ee97edc882b3913091786d5272d7a050865043fee8990b","last_reissued_at":"2026-06-02T01:03:55.339270Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:55.339270Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SALSA: Speech Aware LLM Adaptation via Learned Steering Activation Vectors","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["eess.AS"],"primary_cat":"cs.CL","authors_text":"Argyrios Gerogiannis, Chang D. Yoo, Haolong Zheng, Julia Hockenmaier, Mark A. Hasegawa-Johnson, Yekaterina Yegorova","submitted_at":"2026-05-30T00:54:53Z","abstract_excerpt":"Speech-aware large language models often generalize poorly to out-of-domain settings. We propose SALSA (Speech-Aware LLM Adaptation via Learned Steering Activations), a lightweight adaptation method that learns layer-wise steering vectors. Unlike commonly used steering approaches that rely on contrastive activation differences, SALSA directly optimizes steering vectors using a supervised objective. Across children's speech, multilingual speech, and Mandarin-English code-switching benchmarks, SALSA substantially improves performance over zero-shot inference and speech in-context learning baseli"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.00460","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/2606.00460/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":"2606.00460","created_at":"2026-06-02T01:03:55.339322+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.00460v1","created_at":"2026-06-02T01:03:55.339322+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.00460","created_at":"2026-06-02T01:03:55.339322+00:00"},{"alias_kind":"pith_short_12","alias_value":"YHWPJFFREDHJ","created_at":"2026-06-02T01:03:55.339322+00:00"},{"alias_kind":"pith_short_16","alias_value":"YHWPJFFREDHJ7FHO","created_at":"2026-06-02T01:03:55.339322+00:00"},{"alias_kind":"pith_short_8","alias_value":"YHWPJFFR","created_at":"2026-06-02T01:03:55.339322+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/YHWPJFFREDHJ7FHOS7W4RAVTSE","json":"https://pith.science/pith/YHWPJFFREDHJ7FHOS7W4RAVTSE.json","graph_json":"https://pith.science/api/pith-number/YHWPJFFREDHJ7FHOS7W4RAVTSE/graph.json","events_json":"https://pith.science/api/pith-number/YHWPJFFREDHJ7FHOS7W4RAVTSE/events.json","paper":"https://pith.science/paper/YHWPJFFR"},"agent_actions":{"view_html":"https://pith.science/pith/YHWPJFFREDHJ7FHOS7W4RAVTSE","download_json":"https://pith.science/pith/YHWPJFFREDHJ7FHOS7W4RAVTSE.json","view_paper":"https://pith.science/paper/YHWPJFFR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.00460&json=true","fetch_graph":"https://pith.science/api/pith-number/YHWPJFFREDHJ7FHOS7W4RAVTSE/graph.json","fetch_events":"https://pith.science/api/pith-number/YHWPJFFREDHJ7FHOS7W4RAVTSE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YHWPJFFREDHJ7FHOS7W4RAVTSE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YHWPJFFREDHJ7FHOS7W4RAVTSE/action/storage_attestation","attest_author":"https://pith.science/pith/YHWPJFFREDHJ7FHOS7W4RAVTSE/action/author_attestation","sign_citation":"https://pith.science/pith/YHWPJFFREDHJ7FHOS7W4RAVTSE/action/citation_signature","submit_replication":"https://pith.science/pith/YHWPJFFREDHJ7FHOS7W4RAVTSE/action/replication_record"}},"created_at":"2026-06-02T01:03:55.339322+00:00","updated_at":"2026-06-02T01:03:55.339322+00:00"}