{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DB6RKJCSLIM2FRGJ4X75UKG7QW","short_pith_number":"pith:DB6RKJCS","schema_version":"1.0","canonical_sha256":"187d1524525a19a2c4c9e5ffda28df8596348727ac175e66c40c74d8703a802a","source":{"kind":"arxiv","id":"2502.08866","version":1},"attestation_state":"computed","paper":{"title":"BrainWavLM: Fine-tuning Speech Representations with Brain Responses to Language","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aditya R. Vaidya, Alexander G. Huth, Nishitha Vattikonda, Richard J. Antonello","submitted_at":"2025-02-13T00:37:27Z","abstract_excerpt":"Speech encoding models use auditory representations to predict how the human brain responds to spoken language stimuli. Most performant encoding models linearly map the hidden states of artificial neural networks to brain data, but this linear restriction may limit their effectiveness. In this work, we use low-rank adaptation (LoRA) to fine-tune a WavLM-based encoding model end-to-end on a brain encoding objective, producing a model we name BrainWavLM. We show that fine-tuning across all of cortex improves average encoding performance with greater stability than without LoRA. This improvement "},"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":"2502.08866","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T00:37:27Z","cross_cats_sorted":[],"title_canon_sha256":"3eba77aa467d375b8e5e1ddd27c5ba492a205c2f0f705c88c03ce6d0bd9d1e96","abstract_canon_sha256":"e0bc1e91bd2b91d595bf06dc8fd3635a4f33468a3c2f69fb4393462ef0bd872c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:13:48.861106Z","signature_b64":"tTgX6eJuLn2YoUvgK9iD4hbggrPaveCSkcxD3HZ2VBiT6NI3UJ/A/l1eAYLC7tsZ/c7HaIaWxrhweusF1hE1BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"187d1524525a19a2c4c9e5ffda28df8596348727ac175e66c40c74d8703a802a","last_reissued_at":"2026-07-05T10:13:48.860629Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:13:48.860629Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BrainWavLM: Fine-tuning Speech Representations with Brain Responses to Language","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aditya R. Vaidya, Alexander G. Huth, Nishitha Vattikonda, Richard J. Antonello","submitted_at":"2025-02-13T00:37:27Z","abstract_excerpt":"Speech encoding models use auditory representations to predict how the human brain responds to spoken language stimuli. Most performant encoding models linearly map the hidden states of artificial neural networks to brain data, but this linear restriction may limit their effectiveness. In this work, we use low-rank adaptation (LoRA) to fine-tune a WavLM-based encoding model end-to-end on a brain encoding objective, producing a model we name BrainWavLM. We show that fine-tuning across all of cortex improves average encoding performance with greater stability than without LoRA. This improvement "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.08866","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/2502.08866/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":"2502.08866","created_at":"2026-07-05T10:13:48.860687+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.08866v1","created_at":"2026-07-05T10:13:48.860687+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.08866","created_at":"2026-07-05T10:13:48.860687+00:00"},{"alias_kind":"pith_short_12","alias_value":"DB6RKJCSLIM2","created_at":"2026-07-05T10:13:48.860687+00:00"},{"alias_kind":"pith_short_16","alias_value":"DB6RKJCSLIM2FRGJ","created_at":"2026-07-05T10:13:48.860687+00:00"},{"alias_kind":"pith_short_8","alias_value":"DB6RKJCS","created_at":"2026-07-05T10:13:48.860687+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.08277","citing_title":"Task-conditioned probing of instruction-tuned multimodal LLMs: Region-specific brain alignment patterns under naturalistic stimuli","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09243","citing_title":"How Much is Brain Data Worth for Machine Learning?","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04326","citing_title":"A foundation model of vision, audition, and language for in-silico neuroscience","ref_index":94,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW","json":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW.json","graph_json":"https://pith.science/api/pith-number/DB6RKJCSLIM2FRGJ4X75UKG7QW/graph.json","events_json":"https://pith.science/api/pith-number/DB6RKJCSLIM2FRGJ4X75UKG7QW/events.json","paper":"https://pith.science/paper/DB6RKJCS"},"agent_actions":{"view_html":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW","download_json":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW.json","view_paper":"https://pith.science/paper/DB6RKJCS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.08866&json=true","fetch_graph":"https://pith.science/api/pith-number/DB6RKJCSLIM2FRGJ4X75UKG7QW/graph.json","fetch_events":"https://pith.science/api/pith-number/DB6RKJCSLIM2FRGJ4X75UKG7QW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW/action/storage_attestation","attest_author":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW/action/author_attestation","sign_citation":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW/action/citation_signature","submit_replication":"https://pith.science/pith/DB6RKJCSLIM2FRGJ4X75UKG7QW/action/replication_record"}},"created_at":"2026-07-05T10:13:48.860687+00:00","updated_at":"2026-07-05T10:13:48.860687+00:00"}