{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:4TUPT7L2UHSC7OBKE3IVK4K666","short_pith_number":"pith:4TUPT7L2","schema_version":"1.0","canonical_sha256":"e4e8f9fd7aa1e42fb82a26d155715ef7a2aa09cb810bfc147c4f8abb9e920247","source":{"kind":"arxiv","id":"1808.07982","version":1},"attestation_state":"computed","paper":{"title":"Proximal Policy Optimization and its Dynamic Version for Sequence Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Hung-yi Lee, Jinzhi Zhang, Yi-Lin Tuan, Yujia Li","submitted_at":"2018-08-24T02:14:43Z","abstract_excerpt":"In sequence generation task, many works use policy gradient for model optimization to tackle the intractable backpropagation issue when maximizing the non-differentiable evaluation metrics or fooling the discriminator in adversarial learning. In this paper, we replace policy gradient with proximal policy optimization (PPO), which is a proved more efficient reinforcement learning algorithm, and propose a dynamic approach for PPO (PPO-dynamic). We demonstrate the efficacy of PPO and PPO-dynamic on conditional sequence generation tasks including synthetic experiment and chit-chat chatbot. The res"},"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":"1808.07982","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-08-24T02:14:43Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"61224080f1a1ae947cbab93c00a7d29a1eeb081a7fd0c137278c6a65f9250e19","abstract_canon_sha256":"249d361ef9d861ce4282e24080b48cbd4c9d303cd5ebe881c142186d98cba9fe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:07:22.870860Z","signature_b64":"BbtJ3+avqsHijh0ujOvePbWvL7VPMhpZ6Z+rsFlhSo2C21OUicsesDLuoFTfnIE6f9uBpJr4h892INQYrF5AAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e4e8f9fd7aa1e42fb82a26d155715ef7a2aa09cb810bfc147c4f8abb9e920247","last_reissued_at":"2026-05-18T00:07:22.870349Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:07:22.870349Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Proximal Policy Optimization and its Dynamic Version for Sequence Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Hung-yi Lee, Jinzhi Zhang, Yi-Lin Tuan, Yujia Li","submitted_at":"2018-08-24T02:14:43Z","abstract_excerpt":"In sequence generation task, many works use policy gradient for model optimization to tackle the intractable backpropagation issue when maximizing the non-differentiable evaluation metrics or fooling the discriminator in adversarial learning. In this paper, we replace policy gradient with proximal policy optimization (PPO), which is a proved more efficient reinforcement learning algorithm, and propose a dynamic approach for PPO (PPO-dynamic). We demonstrate the efficacy of PPO and PPO-dynamic on conditional sequence generation tasks including synthetic experiment and chit-chat chatbot. The res"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.07982","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":"1808.07982","created_at":"2026-05-18T00:07:22.870424+00:00"},{"alias_kind":"arxiv_version","alias_value":"1808.07982v1","created_at":"2026-05-18T00:07:22.870424+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.07982","created_at":"2026-05-18T00:07:22.870424+00:00"},{"alias_kind":"pith_short_12","alias_value":"4TUPT7L2UHSC","created_at":"2026-05-18T12:32:05.422762+00:00"},{"alias_kind":"pith_short_16","alias_value":"4TUPT7L2UHSC7OBK","created_at":"2026-05-18T12:32:05.422762+00:00"},{"alias_kind":"pith_short_8","alias_value":"4TUPT7L2","created_at":"2026-05-18T12:32:05.422762+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"1908.05604","citing_title":"Generative Question Refinement with Deep Reinforcement Learning in Retrieval-based QA System","ref_index":34,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666","json":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666.json","graph_json":"https://pith.science/api/pith-number/4TUPT7L2UHSC7OBKE3IVK4K666/graph.json","events_json":"https://pith.science/api/pith-number/4TUPT7L2UHSC7OBKE3IVK4K666/events.json","paper":"https://pith.science/paper/4TUPT7L2"},"agent_actions":{"view_html":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666","download_json":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666.json","view_paper":"https://pith.science/paper/4TUPT7L2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1808.07982&json=true","fetch_graph":"https://pith.science/api/pith-number/4TUPT7L2UHSC7OBKE3IVK4K666/graph.json","fetch_events":"https://pith.science/api/pith-number/4TUPT7L2UHSC7OBKE3IVK4K666/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666/action/storage_attestation","attest_author":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666/action/author_attestation","sign_citation":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666/action/citation_signature","submit_replication":"https://pith.science/pith/4TUPT7L2UHSC7OBKE3IVK4K666/action/replication_record"}},"created_at":"2026-05-18T00:07:22.870424+00:00","updated_at":"2026-05-18T00:07:22.870424+00:00"}