{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:5ZOCOI2ER2IVBGP5JOEIN24QIG","short_pith_number":"pith:5ZOCOI2E","schema_version":"1.0","canonical_sha256":"ee5c2723448e915099fd4b8886eb9041abcf55cfc7efd099c788d1c03d5618d5","source":{"kind":"arxiv","id":"2605.14156","version":1},"attestation_state":"computed","paper":{"title":"Uncovering Trajectory and Topological Signatures in Multimodal Pediatric Sleep Embeddings","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Augmenting multimodal pediatric sleep embeddings with geometric, topological, and clinical features yields complementary gains in predicting breathing events and arousals.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Harlin Lee, Scott Ye","submitted_at":"2026-05-13T22:11:28Z","abstract_excerpt":"While generative models have shown promise in pediatric sleep analysis, the latent structure of their multimodal embeddings remains poorly understood. This work investigates session-wide diagnostic information contained in the sequences of 30-second pediatric PSG epochs embedded by a multimodal masked autoencoder. We test whether augmenting embeddings with PHATE-derived per-epoch coordinates and whole-night movement descriptors, persistent homology summaries of the embedding cloud, and EHR yields task-relevant signals. Simple linear and MLP models, chosen for interpretability rather than state"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.14156","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T22:11:28Z","cross_cats_sorted":[],"title_canon_sha256":"9927227da91f02481ca93903bd8db80d5218594b35f3c3347057b15cc45411f3","abstract_canon_sha256":"48af25b02f32bcf29f496d1b972e5afd3e23ebcd36999c0d3cd197e111602913"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:11.532671Z","signature_b64":"HefRPVikIaqG29xBFlQxZuXGd5juhjrwoGDgsR+N8X25nFytR1Gffj+mvlxVb/vhbzOnAzlVRDcFj1MLPXj2DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ee5c2723448e915099fd4b8886eb9041abcf55cfc7efd099c788d1c03d5618d5","last_reissued_at":"2026-05-17T23:39:11.532013Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:11.532013Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Uncovering Trajectory and Topological Signatures in Multimodal Pediatric Sleep Embeddings","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Augmenting multimodal pediatric sleep embeddings with geometric, topological, and clinical features yields complementary gains in predicting breathing events and arousals.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Harlin Lee, Scott Ye","submitted_at":"2026-05-13T22:11:28Z","abstract_excerpt":"While generative models have shown promise in pediatric sleep analysis, the latent structure of their multimodal embeddings remains poorly understood. This work investigates session-wide diagnostic information contained in the sequences of 30-second pediatric PSG epochs embedded by a multimodal masked autoencoder. We test whether augmenting embeddings with PHATE-derived per-epoch coordinates and whole-night movement descriptors, persistent homology summaries of the embedding cloud, and EHR yields task-relevant signals. Simple linear and MLP models, chosen for interpretability rather than state"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Simple linear and MLP models show that geometric, topological, and clinical features each provide complementary gains, with AUPRC improving from 0.26 to 0.34 for desaturation, 0.31 to 0.48 for EEG arousal, 0.09 to 0.22 for hypopnea, and 0.05 to 0.14 for apnea. The full fusion model yields the best calibration across all four binary tasks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the reported performance gains arise from truly complementary signals in the added features rather than from overfitting, post-hoc feature selection, or dataset-specific artifacts in the imbalanced pediatric PSG data.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Augmenting multimodal pediatric sleep embeddings with PHATE trajectories, persistent homology, movement descriptors, and EHR improves AUPRC and calibration for predicting desaturation, EEG arousal, hypopnea, and apnea.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Augmenting multimodal pediatric sleep embeddings with geometric, topological, and clinical features yields complementary gains in predicting breathing events and arousals.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"fb6f1c66f09b80fdb208b546037da2068c98941d3ff9ebf23148c3de968feb2b"},"source":{"id":"2605.14156","kind":"arxiv","version":1},"verdict":{"id":"3f4f7257-2061-4908-a995-5261d0a573d7","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T04:43:08.653346Z","strongest_claim":"Simple linear and MLP models show that geometric, topological, and clinical features each provide complementary gains, with AUPRC improving from 0.26 to 0.34 for desaturation, 0.31 to 0.48 for EEG arousal, 0.09 to 0.22 for hypopnea, and 0.05 to 0.14 for apnea. The full fusion model yields the best calibration across all four binary tasks.","one_line_summary":"Augmenting multimodal pediatric sleep embeddings with PHATE trajectories, persistent homology, movement descriptors, and EHR improves AUPRC and calibration for predicting desaturation, EEG arousal, hypopnea, and apnea.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the reported performance gains arise from truly complementary signals in the added features rather than from overfitting, post-hoc feature selection, or dataset-specific artifacts in the imbalanced pediatric PSG data.","pith_extraction_headline":"Augmenting multimodal pediatric sleep embeddings with geometric, topological, and clinical features yields complementary gains in predicting breathing events and arousals."},"references":{"count":41,"sample":[{"doi":"","year":2004,"title":"Guillaume Tauzin and Umberto Lupo and Lewis Tunstall and Julian Burella Pérez and Matteo Caorsi and Anibal Medina-Mardones and Alberto Dassatti and Kathryn Hess , year=. giotto-tda:. 2004.02551 , arch","work_id":"0b3fa533-a37d-4de1-8ea1-bc1b13c38e2b","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2000,"title":"Goldberger, Ary L and Amaral, Luis AN and Glass, Leon and Hausdorff, Jeffrey M and Ivanov, Plamen Ch and Mark, Roger G and Mietus, Joseph E and Moody, George B and Peng, Chung-Kang and Stanley, H Euge","work_id":"31f791df-0742-4890-b983-192e7bea96c8","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"Zhang, Guo-Qiang and Cui, Licong and Mueller, Remo and Tao, Shiqiang and Kim, Matthew and Rueschman, Michael and Mariani, Sara and Mobley, Daniel and Redline, Susan , journal=. 2018 , publisher=","work_id":"c960538c-35cf-4772-beb8-e08085f41c68","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"2024 , volume =","work_id":"db960046-13fc-42e6-b48a-e3c6d99639a8","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , month =","work_id":"b48e64c9-dfd9-4b12-8d7b-c0058e5a3f30","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":41,"snapshot_sha256":"a8535aafda457b3f8f5c58459fca75b1c432ea5a6a120c84aa00e3b92e8ac079","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"1119f7bb70a67d26d672aabdf763fa4f50d250a51e6a58022101ae684de4b4ce"},"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":"2605.14156","created_at":"2026-05-17T23:39:11.532104+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.14156v1","created_at":"2026-05-17T23:39:11.532104+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.14156","created_at":"2026-05-17T23:39:11.532104+00:00"},{"alias_kind":"pith_short_12","alias_value":"5ZOCOI2ER2IV","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"5ZOCOI2ER2IVBGP5","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"5ZOCOI2E","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG","json":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG.json","graph_json":"https://pith.science/api/pith-number/5ZOCOI2ER2IVBGP5JOEIN24QIG/graph.json","events_json":"https://pith.science/api/pith-number/5ZOCOI2ER2IVBGP5JOEIN24QIG/events.json","paper":"https://pith.science/paper/5ZOCOI2E"},"agent_actions":{"view_html":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG","download_json":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG.json","view_paper":"https://pith.science/paper/5ZOCOI2E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.14156&json=true","fetch_graph":"https://pith.science/api/pith-number/5ZOCOI2ER2IVBGP5JOEIN24QIG/graph.json","fetch_events":"https://pith.science/api/pith-number/5ZOCOI2ER2IVBGP5JOEIN24QIG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG/action/storage_attestation","attest_author":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG/action/author_attestation","sign_citation":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG/action/citation_signature","submit_replication":"https://pith.science/pith/5ZOCOI2ER2IVBGP5JOEIN24QIG/action/replication_record"}},"created_at":"2026-05-17T23:39:11.532104+00:00","updated_at":"2026-05-17T23:39:11.532104+00:00"}