{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:FK6JQG4LG2GUYBMAXFTCWK6YHJ","short_pith_number":"pith:FK6JQG4L","schema_version":"1.0","canonical_sha256":"2abc981b8b368d4c0580b9662b2bd83a6297e9d3cdaff6646ea61c0aa7db944b","source":{"kind":"arxiv","id":"1708.08863","version":2},"attestation_state":"computed","paper":{"title":"Gradual Learning of Recurrent Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IT","cs.LG","math.IT"],"primary_cat":"stat.ML","authors_text":"Gal Rattner, Haim Permuter, Ziv Aharoni","submitted_at":"2017-08-29T16:18:44Z","abstract_excerpt":"Recurrent Neural Networks (RNNs) achieve state-of-the-art results in many sequence-to-sequence modeling tasks. However, RNNs are difficult to train and tend to suffer from overfitting. Motivated by the Data Processing Inequality (DPI), we formulate the multi-layered network as a Markov chain, introducing a training method that comprises training the network gradually and using layer-wise gradient clipping. We found that applying our methods, combined with previously introduced regularization and optimization methods, resulted in improvements in state-of-the-art architectures operating in langu"},"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":"1708.08863","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2017-08-29T16:18:44Z","cross_cats_sorted":["cs.IT","cs.LG","math.IT"],"title_canon_sha256":"31351208afc3e47aba94753592d9be24c171df0763ca6537fde766eae8dcc9fe","abstract_canon_sha256":"2b0fb30f9887a49199a78dcd0b1c138b89fff6966f8352a932e17c4eacfc7415"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:15:10.261357Z","signature_b64":"ajRxWNqrKD2En4/ka1btLgYNXh7q3EffYemSLaif9SqAjY798J1ivb3R2ku0f+Umf4G37LFJwsH9NIktkFVpCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2abc981b8b368d4c0580b9662b2bd83a6297e9d3cdaff6646ea61c0aa7db944b","last_reissued_at":"2026-05-18T00:15:10.260602Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:15:10.260602Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradual Learning of Recurrent Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IT","cs.LG","math.IT"],"primary_cat":"stat.ML","authors_text":"Gal Rattner, Haim Permuter, Ziv Aharoni","submitted_at":"2017-08-29T16:18:44Z","abstract_excerpt":"Recurrent Neural Networks (RNNs) achieve state-of-the-art results in many sequence-to-sequence modeling tasks. However, RNNs are difficult to train and tend to suffer from overfitting. Motivated by the Data Processing Inequality (DPI), we formulate the multi-layered network as a Markov chain, introducing a training method that comprises training the network gradually and using layer-wise gradient clipping. We found that applying our methods, combined with previously introduced regularization and optimization methods, resulted in improvements in state-of-the-art architectures operating in langu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1708.08863","kind":"arxiv","version":2},"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":"1708.08863","created_at":"2026-05-18T00:15:10.260729+00:00"},{"alias_kind":"arxiv_version","alias_value":"1708.08863v2","created_at":"2026-05-18T00:15:10.260729+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1708.08863","created_at":"2026-05-18T00:15:10.260729+00:00"},{"alias_kind":"pith_short_12","alias_value":"FK6JQG4LG2GU","created_at":"2026-05-18T12:31:15.632608+00:00"},{"alias_kind":"pith_short_16","alias_value":"FK6JQG4LG2GUYBMA","created_at":"2026-05-18T12:31:15.632608+00:00"},{"alias_kind":"pith_short_8","alias_value":"FK6JQG4L","created_at":"2026-05-18T12:31:15.632608+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/FK6JQG4LG2GUYBMAXFTCWK6YHJ","json":"https://pith.science/pith/FK6JQG4LG2GUYBMAXFTCWK6YHJ.json","graph_json":"https://pith.science/api/pith-number/FK6JQG4LG2GUYBMAXFTCWK6YHJ/graph.json","events_json":"https://pith.science/api/pith-number/FK6JQG4LG2GUYBMAXFTCWK6YHJ/events.json","paper":"https://pith.science/paper/FK6JQG4L"},"agent_actions":{"view_html":"https://pith.science/pith/FK6JQG4LG2GUYBMAXFTCWK6YHJ","download_json":"https://pith.science/pith/FK6JQG4LG2GUYBMAXFTCWK6YHJ.json","view_paper":"https://pith.science/paper/FK6JQG4L","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1708.08863&json=true","fetch_graph":"https://pith.science/api/pith-number/FK6JQG4LG2GUYBMAXFTCWK6YHJ/graph.json","fetch_events":"https://pith.science/api/pith-number/FK6JQG4LG2GUYBMAXFTCWK6YHJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FK6JQG4LG2GUYBMAXFTCWK6YHJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FK6JQG4LG2GUYBMAXFTCWK6YHJ/action/storage_attestation","attest_author":"https://pith.science/pith/FK6JQG4LG2GUYBMAXFTCWK6YHJ/action/author_attestation","sign_citation":"https://pith.science/pith/FK6JQG4LG2GUYBMAXFTCWK6YHJ/action/citation_signature","submit_replication":"https://pith.science/pith/FK6JQG4LG2GUYBMAXFTCWK6YHJ/action/replication_record"}},"created_at":"2026-05-18T00:15:10.260729+00:00","updated_at":"2026-05-18T00:15:10.260729+00:00"}