{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:AGTUNUJLSEPLU4ZWANTYVHUZMS","short_pith_number":"pith:AGTUNUJL","schema_version":"1.0","canonical_sha256":"01a746d12b911eba733603678a9e996483edb81f28324d6fe7a57d3986ad416f","source":{"kind":"arxiv","id":"2606.06479","version":1},"attestation_state":"computed","paper":{"title":"Pretraining Recurrent Networks without Recurrence","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Akarsh Kumar, Phillip Isola","submitted_at":"2026-06-04T17:57:33Z","abstract_excerpt":"Training recurrent neural networks (RNNs) requires assigning credit across long sequences of computations. Standard backpropagation through time (BPTT) addresses this problem poorly: it is sequential in time, limiting parallelism, and suffers from vanishing or exploding gradients, making long-range associations difficult to learn. We propose Supervised Memory Training (SMT), a method for training nonlinear RNNs that sidesteps recurrent credit propagation entirely by reducing RNN training to supervised learning on one-step memory transition labels $(m_t, x_{t+1}) \\rightarrow m_{t+1}$. SMT acqui"},"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":"2606.06479","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-04T17:57:33Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"177263f3e7524acbfff044dbd6a792ea6f86761506065f7a683273e3f3bcdbe2","abstract_canon_sha256":"8e6c3fc79854faffed1b86bd95cff6aa91cb1213314bf7cf2acc443d69bde577"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:15:46.774902Z","signature_b64":"oZ/0SWN/nQtzqIx1mRdbEBlvimb1s60zV8q05gkoRdxMAIGcvNY69suUNe2cJAVzeC+eoJU8AUZerSZkJntrAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"01a746d12b911eba733603678a9e996483edb81f28324d6fe7a57d3986ad416f","last_reissued_at":"2026-06-05T01:15:46.774499Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:15:46.774499Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pretraining Recurrent Networks without Recurrence","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Akarsh Kumar, Phillip Isola","submitted_at":"2026-06-04T17:57:33Z","abstract_excerpt":"Training recurrent neural networks (RNNs) requires assigning credit across long sequences of computations. Standard backpropagation through time (BPTT) addresses this problem poorly: it is sequential in time, limiting parallelism, and suffers from vanishing or exploding gradients, making long-range associations difficult to learn. We propose Supervised Memory Training (SMT), a method for training nonlinear RNNs that sidesteps recurrent credit propagation entirely by reducing RNN training to supervised learning on one-step memory transition labels $(m_t, x_{t+1}) \\rightarrow m_{t+1}$. SMT acqui"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.06479","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/2606.06479/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":"2606.06479","created_at":"2026-06-05T01:15:46.774564+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.06479v1","created_at":"2026-06-05T01:15:46.774564+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.06479","created_at":"2026-06-05T01:15:46.774564+00:00"},{"alias_kind":"pith_short_12","alias_value":"AGTUNUJLSEPL","created_at":"2026-06-05T01:15:46.774564+00:00"},{"alias_kind":"pith_short_16","alias_value":"AGTUNUJLSEPLU4ZW","created_at":"2026-06-05T01:15:46.774564+00:00"},{"alias_kind":"pith_short_8","alias_value":"AGTUNUJL","created_at":"2026-06-05T01:15:46.774564+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/AGTUNUJLSEPLU4ZWANTYVHUZMS","json":"https://pith.science/pith/AGTUNUJLSEPLU4ZWANTYVHUZMS.json","graph_json":"https://pith.science/api/pith-number/AGTUNUJLSEPLU4ZWANTYVHUZMS/graph.json","events_json":"https://pith.science/api/pith-number/AGTUNUJLSEPLU4ZWANTYVHUZMS/events.json","paper":"https://pith.science/paper/AGTUNUJL"},"agent_actions":{"view_html":"https://pith.science/pith/AGTUNUJLSEPLU4ZWANTYVHUZMS","download_json":"https://pith.science/pith/AGTUNUJLSEPLU4ZWANTYVHUZMS.json","view_paper":"https://pith.science/paper/AGTUNUJL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.06479&json=true","fetch_graph":"https://pith.science/api/pith-number/AGTUNUJLSEPLU4ZWANTYVHUZMS/graph.json","fetch_events":"https://pith.science/api/pith-number/AGTUNUJLSEPLU4ZWANTYVHUZMS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AGTUNUJLSEPLU4ZWANTYVHUZMS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AGTUNUJLSEPLU4ZWANTYVHUZMS/action/storage_attestation","attest_author":"https://pith.science/pith/AGTUNUJLSEPLU4ZWANTYVHUZMS/action/author_attestation","sign_citation":"https://pith.science/pith/AGTUNUJLSEPLU4ZWANTYVHUZMS/action/citation_signature","submit_replication":"https://pith.science/pith/AGTUNUJLSEPLU4ZWANTYVHUZMS/action/replication_record"}},"created_at":"2026-06-05T01:15:46.774564+00:00","updated_at":"2026-06-05T01:15:46.774564+00:00"}