{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:TU72MQY7DA64KPAO5YBSQ6T7G2","short_pith_number":"pith:TU72MQY7","schema_version":"1.0","canonical_sha256":"9d3fa6431f183dc53c0eee03287a7f36ab0cc14ce26f3b10464c5f182f2a0eab","source":{"kind":"arxiv","id":"2607.27482","version":1},"attestation_state":"computed","paper":{"title":"Latent States in Neural Networks: Recovering the Temporal Structure of Drifting Data from Model Weights","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Kevin Guan","submitted_at":"2026-07-29T21:51:29Z","abstract_excerpt":"A temporally drifting data stream may pass through discrete regimes rather than changing continuously. We ask whether such regimes are recoverable from the weights of models trained on the stream, using a hidden Markov model (HMM) fit to the chronologically ordered trajectory of those weights. We study this question in two domains known to drift over time: multimodal misinformation detection, using the Fakeddit dataset; and sentiment analysis, using the Yelp dataset. We train classifiers on consecutive temporal windows and fit an HMM to the trajectory of their aligned weights, recovering laten"},"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":"2607.27482","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-07-29T21:51:29Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"a8e969a7ee27f675d5b2a0970ceada2f8c1c99ef02579621a9e0c73d4e537b9a","abstract_canon_sha256":"c0e516d55f40af0391eea512ea354b82ca5cb468b478bb20a6c2ba26130b1b4b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9d3fa6431f183dc53c0eee03287a7f36ab0cc14ce26f3b10464c5f182f2a0eab","last_reissued_at":"2026-07-31T01:24:02.568828Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-31T01:24:02.568828Z"},"graph_snapshot":{"paper":{"title":"Latent States in Neural Networks: Recovering the Temporal Structure of Drifting Data from Model Weights","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Kevin Guan","submitted_at":"2026-07-29T21:51:29Z","abstract_excerpt":"A temporally drifting data stream may pass through discrete regimes rather than changing continuously. We ask whether such regimes are recoverable from the weights of models trained on the stream, using a hidden Markov model (HMM) fit to the chronologically ordered trajectory of those weights. We study this question in two domains known to drift over time: multimodal misinformation detection, using the Fakeddit dataset; and sentiment analysis, using the Yelp dataset. We train classifiers on consecutive temporal windows and fit an HMM to the trajectory of their aligned weights, recovering laten"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.27482","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/2607.27482/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":"2607.27482","created_at":"2026-07-31T01:24:02.576192+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.27482v1","created_at":"2026-07-31T01:24:02.576192+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.27482","created_at":"2026-07-31T01:24:02.576192+00:00"},{"alias_kind":"pith_short_12","alias_value":"TU72MQY7DA64","created_at":"2026-07-31T01:24:02.576192+00:00"},{"alias_kind":"pith_short_16","alias_value":"TU72MQY7DA64KPAO","created_at":"2026-07-31T01:24:02.576192+00:00"},{"alias_kind":"pith_short_8","alias_value":"TU72MQY7","created_at":"2026-07-31T01:24:02.576192+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/TU72MQY7DA64KPAO5YBSQ6T7G2","json":"https://pith.science/pith/TU72MQY7DA64KPAO5YBSQ6T7G2.json","graph_json":"https://pith.science/api/pith-number/TU72MQY7DA64KPAO5YBSQ6T7G2/graph.json","events_json":"https://pith.science/api/pith-number/TU72MQY7DA64KPAO5YBSQ6T7G2/events.json","paper":"https://pith.science/paper/TU72MQY7"},"agent_actions":{"view_html":"https://pith.science/pith/TU72MQY7DA64KPAO5YBSQ6T7G2","download_json":"https://pith.science/pith/TU72MQY7DA64KPAO5YBSQ6T7G2.json","view_paper":"https://pith.science/paper/TU72MQY7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.27482&json=true","fetch_graph":"https://pith.science/api/pith-number/TU72MQY7DA64KPAO5YBSQ6T7G2/graph.json","fetch_events":"https://pith.science/api/pith-number/TU72MQY7DA64KPAO5YBSQ6T7G2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TU72MQY7DA64KPAO5YBSQ6T7G2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TU72MQY7DA64KPAO5YBSQ6T7G2/action/storage_attestation","attest_author":"https://pith.science/pith/TU72MQY7DA64KPAO5YBSQ6T7G2/action/author_attestation","sign_citation":"https://pith.science/pith/TU72MQY7DA64KPAO5YBSQ6T7G2/action/citation_signature","submit_replication":"https://pith.science/pith/TU72MQY7DA64KPAO5YBSQ6T7G2/action/replication_record"}},"created_at":"2026-07-31T01:24:02.576192+00:00","updated_at":"2026-07-31T01:24:02.576192+00:00"}