{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:BGNOTKMUXTYQAPCMEHHVB67OP7","short_pith_number":"pith:BGNOTKMU","schema_version":"1.0","canonical_sha256":"099ae9a994bcf1003c4c21cf50fbee7fe9d8f329cd057549499922aa2448d034","source":{"kind":"arxiv","id":"1708.02182","version":1},"attestation_state":"computed","paper":{"title":"Regularizing and Optimizing LSTM Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.NE"],"primary_cat":"cs.CL","authors_text":"Nitish Shirish Keskar, Richard Socher, Stephen Merity","submitted_at":"2017-08-07T16:03:44Z","abstract_excerpt":"Recurrent neural networks (RNNs), such as long short-term memory networks (LSTMs), serve as a fundamental building block for many sequence learning tasks, including machine translation, language modeling, and question answering. In this paper, we consider the specific problem of word-level language modeling and investigate strategies for regularizing and optimizing LSTM-based models. We propose the weight-dropped LSTM which uses DropConnect on hidden-to-hidden weights as a form of recurrent regularization. Further, we introduce NT-ASGD, a variant of the averaged stochastic gradient method, whe"},"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":"1708.02182","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2017-08-07T16:03:44Z","cross_cats_sorted":["cs.LG","cs.NE"],"title_canon_sha256":"e7dbc91cdde810fff496da1dc96a14d3ba5daf29bd51cc237904b1455b50a43f","abstract_canon_sha256":"5094005a94c09f55c343af9edd840dbf87328187b3249a8ac9ae21cbfcc61a4f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:38:25.038185Z","signature_b64":"2swWtNvwWSFRvWeEMo3F9qheHYObw/dSFoUC3O3gp9J3G64zJen09fzBPw90sDOzH8xHvVxEbPO+nw+BbgLQCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"099ae9a994bcf1003c4c21cf50fbee7fe9d8f329cd057549499922aa2448d034","last_reissued_at":"2026-05-18T00:38:25.037552Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:38:25.037552Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Regularizing and Optimizing LSTM Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.NE"],"primary_cat":"cs.CL","authors_text":"Nitish Shirish Keskar, Richard Socher, Stephen Merity","submitted_at":"2017-08-07T16:03:44Z","abstract_excerpt":"Recurrent neural networks (RNNs), such as long short-term memory networks (LSTMs), serve as a fundamental building block for many sequence learning tasks, including machine translation, language modeling, and question answering. In this paper, we consider the specific problem of word-level language modeling and investigate strategies for regularizing and optimizing LSTM-based models. We propose the weight-dropped LSTM which uses DropConnect on hidden-to-hidden weights as a form of recurrent regularization. Further, we introduce NT-ASGD, a variant of the averaged stochastic gradient method, whe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1708.02182","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":"1708.02182","created_at":"2026-05-18T00:38:25.037638+00:00"},{"alias_kind":"arxiv_version","alias_value":"1708.02182v1","created_at":"2026-05-18T00:38:25.037638+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1708.02182","created_at":"2026-05-18T00:38:25.037638+00:00"},{"alias_kind":"pith_short_12","alias_value":"BGNOTKMUXTYQ","created_at":"2026-05-18T12:31:08.081275+00:00"},{"alias_kind":"pith_short_16","alias_value":"BGNOTKMUXTYQAPCM","created_at":"2026-05-18T12:31:08.081275+00:00"},{"alias_kind":"pith_short_8","alias_value":"BGNOTKMU","created_at":"2026-05-18T12:31:08.081275+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":5,"sample":[{"citing_arxiv_id":"1907.01195","citing_title":"Kite: Automatic speech recognition for unmanned aerial vehicles","ref_index":40,"is_internal_anchor":true},{"citing_arxiv_id":"1907.03187","citing_title":"Applying a Pre-trained Language Model to Spanish Twitter Humor Prediction","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2510.04008","citing_title":"RACE Attention: A Strictly Linear-Time Attention Layer for Training on Outrageously Large Contexts","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"1909.05858","citing_title":"CTRL: A Conditional Transformer Language Model for Controllable Generation","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"1908.10063","citing_title":"FinBERT: Financial Sentiment Analysis with Pre-trained Language Models","ref_index":22,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7","json":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7.json","graph_json":"https://pith.science/api/pith-number/BGNOTKMUXTYQAPCMEHHVB67OP7/graph.json","events_json":"https://pith.science/api/pith-number/BGNOTKMUXTYQAPCMEHHVB67OP7/events.json","paper":"https://pith.science/paper/BGNOTKMU"},"agent_actions":{"view_html":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7","download_json":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7.json","view_paper":"https://pith.science/paper/BGNOTKMU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1708.02182&json=true","fetch_graph":"https://pith.science/api/pith-number/BGNOTKMUXTYQAPCMEHHVB67OP7/graph.json","fetch_events":"https://pith.science/api/pith-number/BGNOTKMUXTYQAPCMEHHVB67OP7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7/action/storage_attestation","attest_author":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7/action/author_attestation","sign_citation":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7/action/citation_signature","submit_replication":"https://pith.science/pith/BGNOTKMUXTYQAPCMEHHVB67OP7/action/replication_record"}},"created_at":"2026-05-18T00:38:25.037638+00:00","updated_at":"2026-05-18T00:38:25.037638+00:00"}