{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:7TH3VAOBZXZODSEQYOFCVSEI26","short_pith_number":"pith:7TH3VAOB","schema_version":"1.0","canonical_sha256":"fccfba81c1cdf2e1c890c38a2ac888d7bd2f7c1c04ab3f25806fda25cffaf246","source":{"kind":"arxiv","id":"2405.16158","version":3},"attestation_state":"computed","paper":{"title":"Bigger, Regularized, Optimistic: scaling for compute and sample-efficient continuous control","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Krzysztof Jankowski, Marek Cygan, Mateusz Ostaszewski, Michal Nauman, Piotr Mi{\\l}o\\'s","submitted_at":"2024-05-25T09:53:25Z","abstract_excerpt":"Sample efficiency in Reinforcement Learning (RL) has traditionally been driven by algorithmic enhancements. In this work, we demonstrate that scaling can also lead to substantial improvements. We conduct a thorough investigation into the interplay of scaling model capacity and domain-specific RL enhancements. These empirical findings inform the design choices underlying our proposed BRO (Bigger, Regularized, Optimistic) algorithm. The key innovation behind BRO is that strong regularization allows for effective scaling of the critic networks, which, paired with optimistic exploration, leads to "},"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":"2405.16158","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-25T09:53:25Z","cross_cats_sorted":[],"title_canon_sha256":"88feb6048c768ae63552194696fd77723a94e7c0cf6e2cf917701d49fa50596b","abstract_canon_sha256":"fca7406018097b979d486f5f1f78b0f40afb2692444e743752b00bfaea45713c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:43:29.770232Z","signature_b64":"THNLYKGXofdRaMmyIiQpFkoMt0QqonJ93phfmGT2e38m50Ml8z5ZAniZI/HtmZTVx7Y8j/Xb0K1Fe8Cl42zrCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fccfba81c1cdf2e1c890c38a2ac888d7bd2f7c1c04ab3f25806fda25cffaf246","last_reissued_at":"2026-07-05T09:43:29.769725Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:43:29.769725Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bigger, Regularized, Optimistic: scaling for compute and sample-efficient continuous control","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Krzysztof Jankowski, Marek Cygan, Mateusz Ostaszewski, Michal Nauman, Piotr Mi{\\l}o\\'s","submitted_at":"2024-05-25T09:53:25Z","abstract_excerpt":"Sample efficiency in Reinforcement Learning (RL) has traditionally been driven by algorithmic enhancements. In this work, we demonstrate that scaling can also lead to substantial improvements. We conduct a thorough investigation into the interplay of scaling model capacity and domain-specific RL enhancements. These empirical findings inform the design choices underlying our proposed BRO (Bigger, Regularized, Optimistic) algorithm. The key innovation behind BRO is that strong regularization allows for effective scaling of the critic networks, which, paired with optimistic exploration, leads to "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.16158","kind":"arxiv","version":3},"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/2405.16158/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":"2405.16158","created_at":"2026-07-05T09:43:29.769792+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.16158v3","created_at":"2026-07-05T09:43:29.769792+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.16158","created_at":"2026-07-05T09:43:29.769792+00:00"},{"alias_kind":"pith_short_12","alias_value":"7TH3VAOBZXZO","created_at":"2026-07-05T09:43:29.769792+00:00"},{"alias_kind":"pith_short_16","alias_value":"7TH3VAOBZXZODSEQ","created_at":"2026-07-05T09:43:29.769792+00:00"},{"alias_kind":"pith_short_8","alias_value":"7TH3VAOB","created_at":"2026-07-05T09:43:29.769792+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.25477","citing_title":"EXPO-FT: Sample-Efficient Reinforcement Learning Finetuning for Vision-Language-Action Models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2411.04832","citing_title":"Plasticity Loss in Deep Reinforcement Learning: A Survey","ref_index":83,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04539","citing_title":"FlashSAC: Fast and Stable Off-Policy Reinforcement Learning for High-Dimensional Robot Control","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04539","citing_title":"FlashSAC: Fast and Stable Off-Policy Reinforcement Learning for High-Dimensional Robot Control","ref_index":60,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26","json":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26.json","graph_json":"https://pith.science/api/pith-number/7TH3VAOBZXZODSEQYOFCVSEI26/graph.json","events_json":"https://pith.science/api/pith-number/7TH3VAOBZXZODSEQYOFCVSEI26/events.json","paper":"https://pith.science/paper/7TH3VAOB"},"agent_actions":{"view_html":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26","download_json":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26.json","view_paper":"https://pith.science/paper/7TH3VAOB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.16158&json=true","fetch_graph":"https://pith.science/api/pith-number/7TH3VAOBZXZODSEQYOFCVSEI26/graph.json","fetch_events":"https://pith.science/api/pith-number/7TH3VAOBZXZODSEQYOFCVSEI26/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26/action/storage_attestation","attest_author":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26/action/author_attestation","sign_citation":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26/action/citation_signature","submit_replication":"https://pith.science/pith/7TH3VAOBZXZODSEQYOFCVSEI26/action/replication_record"}},"created_at":"2026-07-05T09:43:29.769792+00:00","updated_at":"2026-07-05T09:43:29.769792+00:00"}