{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:MEFYDUGR2L6ZEHUVJCZ6O63CHE","short_pith_number":"pith:MEFYDUGR","schema_version":"1.0","canonical_sha256":"610b81d0d1d2fd921e9548b3e77b62390718091946bca5cfec4a1237943ec293","source":{"kind":"arxiv","id":"1709.09346","version":2},"attestation_state":"computed","paper":{"title":"Cold-Start Reinforcement Learning with Softmax Policy Gradient","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Nan Ding, Radu Soricut","submitted_at":"2017-09-27T06:04:03Z","abstract_excerpt":"Policy-gradient approaches to reinforcement learning have two common and undesirable overhead procedures, namely warm-start training and sample variance reduction. In this paper, we describe a reinforcement learning method based on a softmax value function that requires neither of these procedures. Our method combines the advantages of policy-gradient methods with the efficiency and simplicity of maximum-likelihood approaches. We apply this new cold-start reinforcement learning method in training sequence generation models for structured output prediction problems. Empirical evidence validates"},"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":"1709.09346","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-09-27T06:04:03Z","cross_cats_sorted":[],"title_canon_sha256":"68a7063ae097841732924d19cc12f11bd7d80b1650ad192ce9dd72a37b6f070a","abstract_canon_sha256":"64f16bd801fa75da84b2a8469cf020371ca21730d6a2da4aca40ebb826c31f0a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:32:54.760653Z","signature_b64":"esce+84sdx2OMBspHO5GXR9ST+NwVggDR9TVEMlZqxUTX8EabIo65eft3oXnenCL6Y0a+gQdauOvJirjiPZbBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"610b81d0d1d2fd921e9548b3e77b62390718091946bca5cfec4a1237943ec293","last_reissued_at":"2026-05-18T00:32:54.759928Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:32:54.759928Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Cold-Start Reinforcement Learning with Softmax Policy Gradient","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Nan Ding, Radu Soricut","submitted_at":"2017-09-27T06:04:03Z","abstract_excerpt":"Policy-gradient approaches to reinforcement learning have two common and undesirable overhead procedures, namely warm-start training and sample variance reduction. In this paper, we describe a reinforcement learning method based on a softmax value function that requires neither of these procedures. Our method combines the advantages of policy-gradient methods with the efficiency and simplicity of maximum-likelihood approaches. We apply this new cold-start reinforcement learning method in training sequence generation models for structured output prediction problems. Empirical evidence validates"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1709.09346","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":"1709.09346","created_at":"2026-05-18T00:32:54.760044+00:00"},{"alias_kind":"arxiv_version","alias_value":"1709.09346v2","created_at":"2026-05-18T00:32:54.760044+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1709.09346","created_at":"2026-05-18T00:32:54.760044+00:00"},{"alias_kind":"pith_short_12","alias_value":"MEFYDUGR2L6Z","created_at":"2026-05-18T12:31:31.346846+00:00"},{"alias_kind":"pith_short_16","alias_value":"MEFYDUGR2L6ZEHUV","created_at":"2026-05-18T12:31:31.346846+00:00"},{"alias_kind":"pith_short_8","alias_value":"MEFYDUGR","created_at":"2026-05-18T12:31:31.346846+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/MEFYDUGR2L6ZEHUVJCZ6O63CHE","json":"https://pith.science/pith/MEFYDUGR2L6ZEHUVJCZ6O63CHE.json","graph_json":"https://pith.science/api/pith-number/MEFYDUGR2L6ZEHUVJCZ6O63CHE/graph.json","events_json":"https://pith.science/api/pith-number/MEFYDUGR2L6ZEHUVJCZ6O63CHE/events.json","paper":"https://pith.science/paper/MEFYDUGR"},"agent_actions":{"view_html":"https://pith.science/pith/MEFYDUGR2L6ZEHUVJCZ6O63CHE","download_json":"https://pith.science/pith/MEFYDUGR2L6ZEHUVJCZ6O63CHE.json","view_paper":"https://pith.science/paper/MEFYDUGR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1709.09346&json=true","fetch_graph":"https://pith.science/api/pith-number/MEFYDUGR2L6ZEHUVJCZ6O63CHE/graph.json","fetch_events":"https://pith.science/api/pith-number/MEFYDUGR2L6ZEHUVJCZ6O63CHE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MEFYDUGR2L6ZEHUVJCZ6O63CHE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MEFYDUGR2L6ZEHUVJCZ6O63CHE/action/storage_attestation","attest_author":"https://pith.science/pith/MEFYDUGR2L6ZEHUVJCZ6O63CHE/action/author_attestation","sign_citation":"https://pith.science/pith/MEFYDUGR2L6ZEHUVJCZ6O63CHE/action/citation_signature","submit_replication":"https://pith.science/pith/MEFYDUGR2L6ZEHUVJCZ6O63CHE/action/replication_record"}},"created_at":"2026-05-18T00:32:54.760044+00:00","updated_at":"2026-05-18T00:32:54.760044+00:00"}