{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:A54I45WM3R4YKUZORYETI7F7AD","short_pith_number":"pith:A54I45WM","schema_version":"1.0","canonical_sha256":"07788e76ccdc7985532e8e09347cbf00d168b240156b54eb0b90d3ff6318628d","source":{"kind":"arxiv","id":"2102.03718","version":1},"attestation_state":"computed","paper":{"title":"An Analysis of Frame-skipping in Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Archit Gupta, Harshith Goka, Kalpesh Krishna, Shivaram Kalyanakrishnan, Siddharth Aravindan, Varun Bhatt, Vihari Piratla, Vishwajeet Bagdawat","submitted_at":"2021-02-07T04:59:09Z","abstract_excerpt":"In the practice of sequential decision making, agents are often designed to sense state at regular intervals of $d$ time steps, $d > 1$, ignoring state information in between sensing steps. While it is clear that this practice can reduce sensing and compute costs, recent results indicate a further benefit. On many Atari console games, reinforcement learning (RL) algorithms deliver substantially better policies when run with $d > 1$ -- in fact with $d$ even as high as $180$. In this paper, we investigate the role of the parameter $d$ in RL; $d$ is called the \"frame-skip\" parameter, since states"},"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":"2102.03718","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-02-07T04:59:09Z","cross_cats_sorted":[],"title_canon_sha256":"937a1033254f20fb5923750ddf5ce64d5f2b95bfe46446d0dd9d83e7cb87de39","abstract_canon_sha256":"d6794db6cbd023e04acc35bd549f251fcf4b2e07a30f8c5463ba6c36df22b589"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:13:26.511113Z","signature_b64":"0kz0G+sCb2mcNakun4OsZOvfFXdIAuohVlP989PhF4gn08ZiVdxTFOlb92yYPWe2KOTUpxFsVOZYmzaAo3lUBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"07788e76ccdc7985532e8e09347cbf00d168b240156b54eb0b90d3ff6318628d","last_reissued_at":"2026-07-05T02:13:26.510690Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:13:26.510690Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"An Analysis of Frame-skipping in Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Archit Gupta, Harshith Goka, Kalpesh Krishna, Shivaram Kalyanakrishnan, Siddharth Aravindan, Varun Bhatt, Vihari Piratla, Vishwajeet Bagdawat","submitted_at":"2021-02-07T04:59:09Z","abstract_excerpt":"In the practice of sequential decision making, agents are often designed to sense state at regular intervals of $d$ time steps, $d > 1$, ignoring state information in between sensing steps. While it is clear that this practice can reduce sensing and compute costs, recent results indicate a further benefit. On many Atari console games, reinforcement learning (RL) algorithms deliver substantially better policies when run with $d > 1$ -- in fact with $d$ even as high as $180$. In this paper, we investigate the role of the parameter $d$ in RL; $d$ is called the \"frame-skip\" parameter, since states"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.03718","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/2102.03718/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":"2102.03718","created_at":"2026-07-05T02:13:26.510746+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.03718v1","created_at":"2026-07-05T02:13:26.510746+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.03718","created_at":"2026-07-05T02:13:26.510746+00:00"},{"alias_kind":"pith_short_12","alias_value":"A54I45WM3R4Y","created_at":"2026-07-05T02:13:26.510746+00:00"},{"alias_kind":"pith_short_16","alias_value":"A54I45WM3R4YKUZO","created_at":"2026-07-05T02:13:26.510746+00:00"},{"alias_kind":"pith_short_8","alias_value":"A54I45WM","created_at":"2026-07-05T02:13:26.510746+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20911","citing_title":"For How Long Should We Be Punching? Learning Action Duration in Fighting Games","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD","json":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD.json","graph_json":"https://pith.science/api/pith-number/A54I45WM3R4YKUZORYETI7F7AD/graph.json","events_json":"https://pith.science/api/pith-number/A54I45WM3R4YKUZORYETI7F7AD/events.json","paper":"https://pith.science/paper/A54I45WM"},"agent_actions":{"view_html":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD","download_json":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD.json","view_paper":"https://pith.science/paper/A54I45WM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.03718&json=true","fetch_graph":"https://pith.science/api/pith-number/A54I45WM3R4YKUZORYETI7F7AD/graph.json","fetch_events":"https://pith.science/api/pith-number/A54I45WM3R4YKUZORYETI7F7AD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD/action/storage_attestation","attest_author":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD/action/author_attestation","sign_citation":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD/action/citation_signature","submit_replication":"https://pith.science/pith/A54I45WM3R4YKUZORYETI7F7AD/action/replication_record"}},"created_at":"2026-07-05T02:13:26.510746+00:00","updated_at":"2026-07-05T02:13:26.510746+00:00"}