{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:MHYEWR5Y35IGYMSFLPOWWWI5ID","short_pith_number":"pith:MHYEWR5Y","schema_version":"1.0","canonical_sha256":"61f04b47b8df506c32455bdd6b591d40f9d3368598b34a56b1a828f34bde2fcb","source":{"kind":"arxiv","id":"2202.11818","version":1},"attestation_state":"computed","paper":{"title":"Consistent Dropout for Policy Gradient Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Matthew Hausknecht, Nolan Wagener","submitted_at":"2022-02-23T23:00:40Z","abstract_excerpt":"Dropout has long been a staple of supervised learning, but is rarely used in reinforcement learning. We analyze why naive application of dropout is problematic for policy-gradient learning algorithms and introduce consistent dropout, a simple technique to address this instability. We demonstrate consistent dropout enables stable training with A2C and PPO in both continuous and discrete action environments across a wide range of dropout probabilities. Finally, we show that consistent dropout enables the online training of complex architectures such as GPT without needing to disable the model's "},"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":"2202.11818","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-02-23T23:00:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"aeb018d062ba36d2cf45786f0a897aa5fc6234a51f500e6cd8e542d312e8afe0","abstract_canon_sha256":"23c04596bdd5fdc966307ae0bb6b50db14d1c8b58e85643bc5443e1612f3c60c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:59:46.440839Z","signature_b64":"hJx3BPTibTuX4QI+8YjnkyFY3+wphy/suF8yz9VstI/TDc0eCGbeH46w1HXQsgTpSWvh8K9LGsxtToeIk7QKBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"61f04b47b8df506c32455bdd6b591d40f9d3368598b34a56b1a828f34bde2fcb","last_reissued_at":"2026-07-05T03:59:46.440424Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:59:46.440424Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Consistent Dropout for Policy Gradient Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Matthew Hausknecht, Nolan Wagener","submitted_at":"2022-02-23T23:00:40Z","abstract_excerpt":"Dropout has long been a staple of supervised learning, but is rarely used in reinforcement learning. We analyze why naive application of dropout is problematic for policy-gradient learning algorithms and introduce consistent dropout, a simple technique to address this instability. We demonstrate consistent dropout enables stable training with A2C and PPO in both continuous and discrete action environments across a wide range of dropout probabilities. Finally, we show that consistent dropout enables the online training of complex architectures such as GPT without needing to disable the model's "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.11818","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2202.11818/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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":"2202.11818","created_at":"2026-07-05T03:59:46.440484+00:00"},{"alias_kind":"arxiv_version","alias_value":"2202.11818v1","created_at":"2026-07-05T03:59:46.440484+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.11818","created_at":"2026-07-05T03:59:46.440484+00:00"},{"alias_kind":"pith_short_12","alias_value":"MHYEWR5Y35IG","created_at":"2026-07-05T03:59:46.440484+00:00"},{"alias_kind":"pith_short_16","alias_value":"MHYEWR5Y35IGYMSF","created_at":"2026-07-05T03:59:46.440484+00:00"},{"alias_kind":"pith_short_8","alias_value":"MHYEWR5Y","created_at":"2026-07-05T03:59:46.440484+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.04039","citing_title":"Generalized Locomotion in Out-of-distribution Conditions with Robust Transformer","ref_index":23,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID","json":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID.json","graph_json":"https://pith.science/api/pith-number/MHYEWR5Y35IGYMSFLPOWWWI5ID/graph.json","events_json":"https://pith.science/api/pith-number/MHYEWR5Y35IGYMSFLPOWWWI5ID/events.json","paper":"https://pith.science/paper/MHYEWR5Y"},"agent_actions":{"view_html":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID","download_json":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID.json","view_paper":"https://pith.science/paper/MHYEWR5Y","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2202.11818&json=true","fetch_graph":"https://pith.science/api/pith-number/MHYEWR5Y35IGYMSFLPOWWWI5ID/graph.json","fetch_events":"https://pith.science/api/pith-number/MHYEWR5Y35IGYMSFLPOWWWI5ID/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID/action/storage_attestation","attest_author":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID/action/author_attestation","sign_citation":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID/action/citation_signature","submit_replication":"https://pith.science/pith/MHYEWR5Y35IGYMSFLPOWWWI5ID/action/replication_record"}},"created_at":"2026-07-05T03:59:46.440484+00:00","updated_at":"2026-07-05T03:59:46.440484+00:00"}