{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YNHUEXDQOPRKK77VZQPYFFTKY7","short_pith_number":"pith:YNHUEXDQ","schema_version":"1.0","canonical_sha256":"c34f425c7073e2a57ff5cc1f82966ac7fcec8c0c355bcb3897a446b310e5eb0c","source":{"kind":"arxiv","id":"2607.01218","version":1},"attestation_state":"computed","paper":{"title":"The State-Prediction Separation Hypothesis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Giovanni Monea, Kiant\\'e Brantley, Nathan Godey, Yoav Artzi","submitted_at":"2026-07-01T17:55:09Z","abstract_excerpt":"Transformers use the same forward computation stream to both predict the next token and store useful state for future token predictions. We formulate the \\emph{state-prediction separation hypothesis}: disentangling the two roles yields better language modeling performance. We design a Transformer variant that uses two computation streams to separate the two functions, and conduct pretraining experiments across various scales. Our experiments show that state-prediction separation consistently offers better data and compute efficiencies, improving validation loss and outperforming standard Trans"},"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":"2607.01218","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-07-01T17:55:09Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"46d24c7fff5881eeac40386f39f83894a5f45ebf65ce0e7c73db67085a756985","abstract_canon_sha256":"7605b9a0ae4df7d9f184d8185f6accadc0a6bda07eaa65074ca2caf837fd0619"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-02T01:18:33.032583Z","signature_b64":"DVhO2n+Xmz3pGleQxsiYxF4nuTEpXYdjiCldpC3zwLNZlCHrZ08fQiMTkgBEe75CTKyBpaSfS4MJYkDHyUH0Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c34f425c7073e2a57ff5cc1f82966ac7fcec8c0c355bcb3897a446b310e5eb0c","last_reissued_at":"2026-07-02T01:18:33.032186Z","signature_status":"signed_v1","first_computed_at":"2026-07-02T01:18:33.032186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The State-Prediction Separation Hypothesis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Giovanni Monea, Kiant\\'e Brantley, Nathan Godey, Yoav Artzi","submitted_at":"2026-07-01T17:55:09Z","abstract_excerpt":"Transformers use the same forward computation stream to both predict the next token and store useful state for future token predictions. We formulate the \\emph{state-prediction separation hypothesis}: disentangling the two roles yields better language modeling performance. We design a Transformer variant that uses two computation streams to separate the two functions, and conduct pretraining experiments across various scales. Our experiments show that state-prediction separation consistently offers better data and compute efficiencies, improving validation loss and outperforming standard Trans"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.01218","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/2607.01218/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":"2607.01218","created_at":"2026-07-02T01:18:33.032248+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.01218v1","created_at":"2026-07-02T01:18:33.032248+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.01218","created_at":"2026-07-02T01:18:33.032248+00:00"},{"alias_kind":"pith_short_12","alias_value":"YNHUEXDQOPRK","created_at":"2026-07-02T01:18:33.032248+00:00"},{"alias_kind":"pith_short_16","alias_value":"YNHUEXDQOPRKK77V","created_at":"2026-07-02T01:18:33.032248+00:00"},{"alias_kind":"pith_short_8","alias_value":"YNHUEXDQ","created_at":"2026-07-02T01:18:33.032248+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/YNHUEXDQOPRKK77VZQPYFFTKY7","json":"https://pith.science/pith/YNHUEXDQOPRKK77VZQPYFFTKY7.json","graph_json":"https://pith.science/api/pith-number/YNHUEXDQOPRKK77VZQPYFFTKY7/graph.json","events_json":"https://pith.science/api/pith-number/YNHUEXDQOPRKK77VZQPYFFTKY7/events.json","paper":"https://pith.science/paper/YNHUEXDQ"},"agent_actions":{"view_html":"https://pith.science/pith/YNHUEXDQOPRKK77VZQPYFFTKY7","download_json":"https://pith.science/pith/YNHUEXDQOPRKK77VZQPYFFTKY7.json","view_paper":"https://pith.science/paper/YNHUEXDQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.01218&json=true","fetch_graph":"https://pith.science/api/pith-number/YNHUEXDQOPRKK77VZQPYFFTKY7/graph.json","fetch_events":"https://pith.science/api/pith-number/YNHUEXDQOPRKK77VZQPYFFTKY7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YNHUEXDQOPRKK77VZQPYFFTKY7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YNHUEXDQOPRKK77VZQPYFFTKY7/action/storage_attestation","attest_author":"https://pith.science/pith/YNHUEXDQOPRKK77VZQPYFFTKY7/action/author_attestation","sign_citation":"https://pith.science/pith/YNHUEXDQOPRKK77VZQPYFFTKY7/action/citation_signature","submit_replication":"https://pith.science/pith/YNHUEXDQOPRKK77VZQPYFFTKY7/action/replication_record"}},"created_at":"2026-07-02T01:18:33.032248+00:00","updated_at":"2026-07-02T01:18:33.032248+00:00"}