{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VT5UIUP4MSENTADGGOR3LHGC6I","short_pith_number":"pith:VT5UIUP4","schema_version":"1.0","canonical_sha256":"acfb4451fc6488d9806633a3b59cc2f21866c2a35aaaaeb4a437fd386effbee3","source":{"kind":"arxiv","id":"2607.22264","version":1},"attestation_state":"computed","paper":{"title":"Autoregressive EHR Foundation Models with Multimodal Inputs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"A. Aldo Faisal, Alfred John Balston, Jinpei Han, Joshua Placidi, Marek Rei, Yuxuan Liu","submitted_at":"2026-07-24T12:59:20Z","abstract_excerpt":"Autoregressive foundation models trained on tokenized electronic health records (EHRs) can support zero-shot clinical prediction, yet most operate on structured event codes alone, and do not incorporate multiple modalities in a principled way. We present a framework for conditioning such models on auxiliary clinical modalities, including ECG waveforms, chest X-ray images, and clinical notes, using modality-specific latent compression and gated cross-attention with temporal alignment. We investigate two key design choices: (1) how to compress long per-modality sequences (e.g., ECG time series) "},"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":"2607.22264","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-07-24T12:59:20Z","cross_cats_sorted":[],"title_canon_sha256":"49b3dad403c2232905eb1d9ea22e7eb85e71e7edeabef96a08a562ad326bde4e","abstract_canon_sha256":"65d0954eb0b2042b0461e6ef64c01c12deab174964650b43c0ee534679288c67"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-27T01:20:56.744860Z","signature_b64":"oGgF0pF1mjwZz8Sr/BtYOk6oHuPLxO6alqBkbAdxqQIqceP+ewMxYMN6JVsYQuQObKHrAuXACdDDFatBL4F5DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"acfb4451fc6488d9806633a3b59cc2f21866c2a35aaaaeb4a437fd386effbee3","last_reissued_at":"2026-07-27T01:20:56.743996Z","signature_status":"signed_v1","first_computed_at":"2026-07-27T01:20:56.743996Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Autoregressive EHR Foundation Models with Multimodal Inputs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"A. Aldo Faisal, Alfred John Balston, Jinpei Han, Joshua Placidi, Marek Rei, Yuxuan Liu","submitted_at":"2026-07-24T12:59:20Z","abstract_excerpt":"Autoregressive foundation models trained on tokenized electronic health records (EHRs) can support zero-shot clinical prediction, yet most operate on structured event codes alone, and do not incorporate multiple modalities in a principled way. We present a framework for conditioning such models on auxiliary clinical modalities, including ECG waveforms, chest X-ray images, and clinical notes, using modality-specific latent compression and gated cross-attention with temporal alignment. We investigate two key design choices: (1) how to compress long per-modality sequences (e.g., ECG time series) "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.22264","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/2607.22264/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":"2607.22264","created_at":"2026-07-27T01:20:56.744448+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.22264v1","created_at":"2026-07-27T01:20:56.744448+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.22264","created_at":"2026-07-27T01:20:56.744448+00:00"},{"alias_kind":"pith_short_12","alias_value":"VT5UIUP4MSEN","created_at":"2026-07-27T01:20:56.744448+00:00"},{"alias_kind":"pith_short_16","alias_value":"VT5UIUP4MSENTADG","created_at":"2026-07-27T01:20:56.744448+00:00"},{"alias_kind":"pith_short_8","alias_value":"VT5UIUP4","created_at":"2026-07-27T01:20:56.744448+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/VT5UIUP4MSENTADGGOR3LHGC6I","json":"https://pith.science/pith/VT5UIUP4MSENTADGGOR3LHGC6I.json","graph_json":"https://pith.science/api/pith-number/VT5UIUP4MSENTADGGOR3LHGC6I/graph.json","events_json":"https://pith.science/api/pith-number/VT5UIUP4MSENTADGGOR3LHGC6I/events.json","paper":"https://pith.science/paper/VT5UIUP4"},"agent_actions":{"view_html":"https://pith.science/pith/VT5UIUP4MSENTADGGOR3LHGC6I","download_json":"https://pith.science/pith/VT5UIUP4MSENTADGGOR3LHGC6I.json","view_paper":"https://pith.science/paper/VT5UIUP4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.22264&json=true","fetch_graph":"https://pith.science/api/pith-number/VT5UIUP4MSENTADGGOR3LHGC6I/graph.json","fetch_events":"https://pith.science/api/pith-number/VT5UIUP4MSENTADGGOR3LHGC6I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VT5UIUP4MSENTADGGOR3LHGC6I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VT5UIUP4MSENTADGGOR3LHGC6I/action/storage_attestation","attest_author":"https://pith.science/pith/VT5UIUP4MSENTADGGOR3LHGC6I/action/author_attestation","sign_citation":"https://pith.science/pith/VT5UIUP4MSENTADGGOR3LHGC6I/action/citation_signature","submit_replication":"https://pith.science/pith/VT5UIUP4MSENTADGGOR3LHGC6I/action/replication_record"}},"created_at":"2026-07-27T01:20:56.744448+00:00","updated_at":"2026-07-27T01:20:56.744448+00:00"}