{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:6T7HNX2JFFNELMAXGSTGRQLNIP","short_pith_number":"pith:6T7HNX2J","schema_version":"1.0","canonical_sha256":"f4fe76df49295a45b01734a668c16d43d9bf5a434ff655f36ecdb8eeaf7659bc","source":{"kind":"arxiv","id":"1804.11188","version":1},"attestation_state":"computed","paper":{"title":"Can recurrent neural networks warp time?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE","stat.ML"],"primary_cat":"cs.LG","authors_text":"Corentin Tallec, Yann Ollivier","submitted_at":"2018-03-23T09:17:35Z","abstract_excerpt":"Successful recurrent models such as long short-term memories (LSTMs) and gated recurrent units (GRUs) use ad hoc gating mechanisms. Empirically these models have been found to improve the learning of medium to long term temporal dependencies and to help with vanishing gradient issues. We prove that learnable gates in a recurrent model formally provide quasi- invariance to general time transformations in the input data. We recover part of the LSTM architecture from a simple axiomatic approach. This result leads to a new way of initializing gate biases in LSTMs and GRUs. Ex- perimentally, this n"},"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":"1804.11188","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-03-23T09:17:35Z","cross_cats_sorted":["cs.NE","stat.ML"],"title_canon_sha256":"ed0389c379dd6608df486bd801488ccabbd82f832143afa0ac7b92aacf9a943a","abstract_canon_sha256":"2ba97eacb30566a57f4046384307f575e5968d421ee8f4198f8c388fba0a44ae"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:17:14.349433Z","signature_b64":"U7Z4kjgtocUZmQYvG+iWI8YDKF5Tc3ESSaneTq8yns/olxCEmtQpVU176yjrDOdlxbFKZLYOYj718USGUDWnDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f4fe76df49295a45b01734a668c16d43d9bf5a434ff655f36ecdb8eeaf7659bc","last_reissued_at":"2026-05-18T00:17:14.348614Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:17:14.348614Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can recurrent neural networks warp time?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE","stat.ML"],"primary_cat":"cs.LG","authors_text":"Corentin Tallec, Yann Ollivier","submitted_at":"2018-03-23T09:17:35Z","abstract_excerpt":"Successful recurrent models such as long short-term memories (LSTMs) and gated recurrent units (GRUs) use ad hoc gating mechanisms. Empirically these models have been found to improve the learning of medium to long term temporal dependencies and to help with vanishing gradient issues. We prove that learnable gates in a recurrent model formally provide quasi- invariance to general time transformations in the input data. We recover part of the LSTM architecture from a simple axiomatic approach. This result leads to a new way of initializing gate biases in LSTMs and GRUs. Ex- perimentally, this n"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1804.11188","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":""},"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":"1804.11188","created_at":"2026-05-18T00:17:14.348731+00:00"},{"alias_kind":"arxiv_version","alias_value":"1804.11188v1","created_at":"2026-05-18T00:17:14.348731+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1804.11188","created_at":"2026-05-18T00:17:14.348731+00:00"},{"alias_kind":"pith_short_12","alias_value":"6T7HNX2JFFNE","created_at":"2026-05-18T12:32:11.075285+00:00"},{"alias_kind":"pith_short_16","alias_value":"6T7HNX2JFFNELMAX","created_at":"2026-05-18T12:32:11.075285+00:00"},{"alias_kind":"pith_short_8","alias_value":"6T7HNX2J","created_at":"2026-05-18T12:32:11.075285+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2602.10385","citing_title":"Capture Timing-Attention of Events in Clinical Time Series","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2104.13478","citing_title":"Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges","ref_index":90,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01134","citing_title":"To Use AI as Dice of Possibilities with Timing Computation","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13857","citing_title":"Mamba Sequence Modeling meets Model Predictive Control","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP","json":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP.json","graph_json":"https://pith.science/api/pith-number/6T7HNX2JFFNELMAXGSTGRQLNIP/graph.json","events_json":"https://pith.science/api/pith-number/6T7HNX2JFFNELMAXGSTGRQLNIP/events.json","paper":"https://pith.science/paper/6T7HNX2J"},"agent_actions":{"view_html":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP","download_json":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP.json","view_paper":"https://pith.science/paper/6T7HNX2J","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1804.11188&json=true","fetch_graph":"https://pith.science/api/pith-number/6T7HNX2JFFNELMAXGSTGRQLNIP/graph.json","fetch_events":"https://pith.science/api/pith-number/6T7HNX2JFFNELMAXGSTGRQLNIP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP/action/storage_attestation","attest_author":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP/action/author_attestation","sign_citation":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP/action/citation_signature","submit_replication":"https://pith.science/pith/6T7HNX2JFFNELMAXGSTGRQLNIP/action/replication_record"}},"created_at":"2026-05-18T00:17:14.348731+00:00","updated_at":"2026-05-18T00:17:14.348731+00:00"}