{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:KANXA7MJHPATK5CC72CCGD4IZ5","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"53dfe2eb4add8b39eff7960bd25d4f9a8f6053c4ea05f9fdd9cb53458f4f63b4","cross_cats_sorted":["cs.AI","cs.DC","physics.comp-ph"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T19:32:32Z","title_canon_sha256":"c3b46719491e04900891462c48b5ecdf47b1afc9a287a877bd088e2267908b41"},"schema_version":"1.0","source":{"id":"2605.12683","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.12683","created_at":"2026-05-18T03:09:49Z"},{"alias_kind":"arxiv_version","alias_value":"2605.12683v1","created_at":"2026-05-18T03:09:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.12683","created_at":"2026-05-18T03:09:49Z"},{"alias_kind":"pith_short_12","alias_value":"KANXA7MJHPAT","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"KANXA7MJHPATK5CC","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"KANXA7MJ","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:b13a8f87b107ed396bb6af60186583db7fc47abd3a92c9d8bce4ee103b511100","target":"graph","created_at":"2026-05-18T03:09:49Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Using GTF-DEER, we investigate the benefits of training on extremely long sequences (T>10^4) for DSR. Our results show that access to such long trajectories significantly improves DSR if the data features long time scales."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the parallel-in-time algorithms, including the new GTF variant, maintain numerical stability and learning effectiveness for general nonlinear dynamics across arbitrary sequence lengths without hidden constraints or post-hoc adjustments."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"GTF-DEER augments the DEER framework with Generalized Teacher Forcing to allow effective parallel training of nonlinear recurrent models on extremely long sequences, improving dynamical systems reconstruction for data with long time scales."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Generalized teacher forcing in the DEER framework enables stable parallel-in-time training of nonlinear recurrent models on sequences longer than 10,000 steps, yielding better reconstruction of dynamical systems with long time scales."}],"snapshot_sha256":"84afee016d95623fd6eb2003c2be86931c822e39ba40ec52698f9d059b1feb7b"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"643427d69d140389bff82a6f9888974e5352f5306de081bd232b59091adc81de"},"paper":{"abstract_excerpt":"Reconstructing nonlinear dynamical systems (DS) from data (DSR) is a fundamental challenge in science and engineering, but it inherently relies on sequential models. Recent breakthroughs for sequential models have produced algorithms that parallelize computation along sequence length $T$, achieving logarithmic time complexity, $\\mathcal{O}(\\log T)$. Since sequence lengths have been practically limited due to the linear runtime complexity $\\mathcal{O}(T)$ of classical backpropagation through time, this opens new avenues for DSR. This paper studies two prominent classes of parallel-in-time algor","authors_text":"Daniel Durstewitz, Florian G\\\"otz, Florian Hess","cross_cats":["cs.AI","cs.DC","physics.comp-ph"],"headline":"Generalized teacher forcing in the DEER framework enables stable parallel-in-time training of nonlinear recurrent models on sequences longer than 10,000 steps, yielding better reconstruction of dynamical systems with long time scales.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T19:32:32Z","title":"Parallel-in-Time Training of Recurrent Neural Networks for Dynamical Systems Reconstruction"},"references":{"count":82,"internal_anchors":3,"resolved_work":82,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Balanced neural ODEs: nonlinear model order reduction and koopman operator approximations","work_id":"113f21b7-6783-4546-b37e-40086b6bdc2c","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Martinez Alvarez, Rare¸ s Ro¸ sca, and Cristian G","work_id":"a927f77c-cf4b-40a8-a1c1-739240b25348","year":2009},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Benjamin Erichson, Vanessa Lin, and Michael W","work_id":"9a4f6106-4060-4a0d-8ebe-2ad199bb71e3","year":2020},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Scheduled sampling for sequence prediction with recurrent neural networks.Advances in neural information processing systems, 28","work_id":"64dc8a72-f476-4512-90fa-5aa8f92cd39d","year":2015},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Y . Bengio, P. Simard, and P. Frasconi. Learning long-term dependencies with gradient descent is difficult.IEEE transactions on neural networks, 5(2):157–166, 1994","work_id":"56539f7f-6c4c-406f-b128-9188a05c6c59","year":1994}],"snapshot_sha256":"a50ea77a158d67152bd0613101555a4c516be7cbc9068e4de6c25ce35e8625b1"},"source":{"id":"2605.12683","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T21:08:48.912559Z","id":"9ec9d8b3-9bf5-4d20-99ee-8374ec8bf2fe","model_set":{"reader":"grok-4.3"},"one_line_summary":"GTF-DEER augments the DEER framework with Generalized Teacher Forcing to allow effective parallel training of nonlinear recurrent models on extremely long sequences, improving dynamical systems reconstruction for data with long time scales.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Generalized teacher forcing in the DEER framework enables stable parallel-in-time training of nonlinear recurrent models on sequences longer than 10,000 steps, yielding better reconstruction of dynamical systems with long time scales.","strongest_claim":"Using GTF-DEER, we investigate the benefits of training on extremely long sequences (T>10^4) for DSR. Our results show that access to such long trajectories significantly improves DSR if the data features long time scales.","weakest_assumption":"That the parallel-in-time algorithms, including the new GTF variant, maintain numerical stability and learning effectiveness for general nonlinear dynamics across arbitrary sequence lengths without hidden constraints or post-hoc adjustments."}},"verdict_id":"9ec9d8b3-9bf5-4d20-99ee-8374ec8bf2fe"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:abb4fad22eabd81f870a73c79d8e61d149a09a8fd84b62dbc1261b7c175530d9","target":"record","created_at":"2026-05-18T03:09:49Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"53dfe2eb4add8b39eff7960bd25d4f9a8f6053c4ea05f9fdd9cb53458f4f63b4","cross_cats_sorted":["cs.AI","cs.DC","physics.comp-ph"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-12T19:32:32Z","title_canon_sha256":"c3b46719491e04900891462c48b5ecdf47b1afc9a287a877bd088e2267908b41"},"schema_version":"1.0","source":{"id":"2605.12683","kind":"arxiv","version":1}},"canonical_sha256":"501b707d893bc1357442fe84230f88cf74aa7e2593a1daefa07b8b6c704bd89b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"501b707d893bc1357442fe84230f88cf74aa7e2593a1daefa07b8b6c704bd89b","first_computed_at":"2026-05-18T03:09:49.971009Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:09:49.971009Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZBvDBbUjTCwzjJ/caK30LB19P8fZsNUUPgjFM4JPDbu+s22trtPYVwOmMRVyLNCGAqH9kH//v9el+5RzlA7WDg==","signature_status":"signed_v1","signed_at":"2026-05-18T03:09:49.971872Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.12683","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:abb4fad22eabd81f870a73c79d8e61d149a09a8fd84b62dbc1261b7c175530d9","sha256:b13a8f87b107ed396bb6af60186583db7fc47abd3a92c9d8bce4ee103b511100"],"state_sha256":"0318e836b738a9ca134a40b6efdbe4f33ee9ffc01a61ad1c16d6c2ff9a91911f"}