{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:6MP5RM3YIJEDGDXIIBX34P45CW","short_pith_number":"pith:6MP5RM3Y","schema_version":"1.0","canonical_sha256":"f31fd8b3784248330ee8406fbe3f9d158bf7474c8afeac38033995399f7d7f6a","source":{"kind":"arxiv","id":"1904.10642","version":1},"attestation_state":"computed","paper":{"title":"Towards Combining On-Off-Policy Methods for Real-World Applications","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Chen-Huan Pi, I-Chen Wu, Kai-Chun Hu, Stone Cheng, Ting Han Wei, Wei-Yuan Ye, Yi-Wei Dai","submitted_at":"2019-04-24T05:07:49Z","abstract_excerpt":"In this paper, we point out a fundamental property of the objective in reinforcement learning, with which we can reformulate the policy gradient objective into a perceptron-like loss function, removing the need to distinguish between on and off policy training. Namely, we posit that it is sufficient to only update a policy $\\pi$ for cases that satisfy the condition $A(\\frac{\\pi}{\\mu}-1)\\leq0$, where $A$ is the advantage, and $\\mu$ is another policy. Furthermore, we show via theoretic derivation that a perceptron-like loss function matches the clipped surrogate objective for PPO. With our new f"},"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":"1904.10642","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-04-24T05:07:49Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"8fc4e5de907cfddc29b7e9a162dd477d3a34480849435884daf49818fc9a7899","abstract_canon_sha256":"2f1d28e1e5fb9cf21852db20838c2a030f440757556e2e27536cc96f737f4ad5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:47:53.146814Z","signature_b64":"f0KRBa3lq7Efa3fIRLi4QXU8FwU+ebtZiygN2Y5NMm9H3oRzA3pZZPmNsUzoakg1IrBn8B6QsaPHEyN4SHi/BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f31fd8b3784248330ee8406fbe3f9d158bf7474c8afeac38033995399f7d7f6a","last_reissued_at":"2026-05-17T23:47:53.146429Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:47:53.146429Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Combining On-Off-Policy Methods for Real-World Applications","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Chen-Huan Pi, I-Chen Wu, Kai-Chun Hu, Stone Cheng, Ting Han Wei, Wei-Yuan Ye, Yi-Wei Dai","submitted_at":"2019-04-24T05:07:49Z","abstract_excerpt":"In this paper, we point out a fundamental property of the objective in reinforcement learning, with which we can reformulate the policy gradient objective into a perceptron-like loss function, removing the need to distinguish between on and off policy training. Namely, we posit that it is sufficient to only update a policy $\\pi$ for cases that satisfy the condition $A(\\frac{\\pi}{\\mu}-1)\\leq0$, where $A$ is the advantage, and $\\mu$ is another policy. Furthermore, we show via theoretic derivation that a perceptron-like loss function matches the clipped surrogate objective for PPO. With our new f"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1904.10642","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":"1904.10642","created_at":"2026-05-17T23:47:53.146487+00:00"},{"alias_kind":"arxiv_version","alias_value":"1904.10642v1","created_at":"2026-05-17T23:47:53.146487+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1904.10642","created_at":"2026-05-17T23:47:53.146487+00:00"},{"alias_kind":"pith_short_12","alias_value":"6MP5RM3YIJED","created_at":"2026-05-18T12:33:10.108867+00:00"},{"alias_kind":"pith_short_16","alias_value":"6MP5RM3YIJEDGDXI","created_at":"2026-05-18T12:33:10.108867+00:00"},{"alias_kind":"pith_short_8","alias_value":"6MP5RM3Y","created_at":"2026-05-18T12:33:10.108867+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/6MP5RM3YIJEDGDXIIBX34P45CW","json":"https://pith.science/pith/6MP5RM3YIJEDGDXIIBX34P45CW.json","graph_json":"https://pith.science/api/pith-number/6MP5RM3YIJEDGDXIIBX34P45CW/graph.json","events_json":"https://pith.science/api/pith-number/6MP5RM3YIJEDGDXIIBX34P45CW/events.json","paper":"https://pith.science/paper/6MP5RM3Y"},"agent_actions":{"view_html":"https://pith.science/pith/6MP5RM3YIJEDGDXIIBX34P45CW","download_json":"https://pith.science/pith/6MP5RM3YIJEDGDXIIBX34P45CW.json","view_paper":"https://pith.science/paper/6MP5RM3Y","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1904.10642&json=true","fetch_graph":"https://pith.science/api/pith-number/6MP5RM3YIJEDGDXIIBX34P45CW/graph.json","fetch_events":"https://pith.science/api/pith-number/6MP5RM3YIJEDGDXIIBX34P45CW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6MP5RM3YIJEDGDXIIBX34P45CW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6MP5RM3YIJEDGDXIIBX34P45CW/action/storage_attestation","attest_author":"https://pith.science/pith/6MP5RM3YIJEDGDXIIBX34P45CW/action/author_attestation","sign_citation":"https://pith.science/pith/6MP5RM3YIJEDGDXIIBX34P45CW/action/citation_signature","submit_replication":"https://pith.science/pith/6MP5RM3YIJEDGDXIIBX34P45CW/action/replication_record"}},"created_at":"2026-05-17T23:47:53.146487+00:00","updated_at":"2026-05-17T23:47:53.146487+00:00"}