{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:EIKKKWLHVJESAH2VXMV7X64LUY","short_pith_number":"pith:EIKKKWLH","schema_version":"1.0","canonical_sha256":"2214a55967aa49201f55bb2bfbfb8ba6245ad77451132f1b867d69c4257e94c7","source":{"kind":"arxiv","id":"2207.07560","version":2},"attestation_state":"computed","paper":{"title":"Skill-based Model-based Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.RO"],"primary_cat":"cs.LG","authors_text":"Joseph J. Lim, Lucy Xiaoyang Shi, Youngwoon Lee","submitted_at":"2022-07-15T16:06:33Z","abstract_excerpt":"Model-based reinforcement learning (RL) is a sample-efficient way of learning complex behaviors by leveraging a learned single-step dynamics model to plan actions in imagination. However, planning every action for long-horizon tasks is not practical, akin to a human planning out every muscle movement. Instead, humans efficiently plan with high-level skills to solve complex tasks. From this intuition, we propose a Skill-based Model-based RL framework (SkiMo) that enables planning in the skill space using a skill dynamics model, which directly predicts the skill outcomes, rather than predicting "},"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":"2207.07560","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-07-15T16:06:33Z","cross_cats_sorted":["cs.AI","cs.RO"],"title_canon_sha256":"64a63dd380e3cff3040507c286f22edbe76df4217bc630292417e08655d590a0","abstract_canon_sha256":"fb3162dd6aecb8a6f934957aa425109df58ea458c24bc463cc44473b1abc7e42"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:24:12.829949Z","signature_b64":"b8DbFHjEjSznCtM7DgUxBwWeEJINB+Ose0NPpoQNgVPLleu+4+iOh6g9r6VFw0nZ5oxNWcfCDSbFZXVrDCk2BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2214a55967aa49201f55bb2bfbfb8ba6245ad77451132f1b867d69c4257e94c7","last_reissued_at":"2026-07-05T05:24:12.829507Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:24:12.829507Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Skill-based Model-based Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.RO"],"primary_cat":"cs.LG","authors_text":"Joseph J. Lim, Lucy Xiaoyang Shi, Youngwoon Lee","submitted_at":"2022-07-15T16:06:33Z","abstract_excerpt":"Model-based reinforcement learning (RL) is a sample-efficient way of learning complex behaviors by leveraging a learned single-step dynamics model to plan actions in imagination. However, planning every action for long-horizon tasks is not practical, akin to a human planning out every muscle movement. Instead, humans efficiently plan with high-level skills to solve complex tasks. From this intuition, we propose a Skill-based Model-based RL framework (SkiMo) that enables planning in the skill space using a skill dynamics model, which directly predicts the skill outcomes, rather than predicting "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.07560","kind":"arxiv","version":2},"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/2207.07560/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":"2207.07560","created_at":"2026-07-05T05:24:12.829565+00:00"},{"alias_kind":"arxiv_version","alias_value":"2207.07560v2","created_at":"2026-07-05T05:24:12.829565+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.07560","created_at":"2026-07-05T05:24:12.829565+00:00"},{"alias_kind":"pith_short_12","alias_value":"EIKKKWLHVJES","created_at":"2026-07-05T05:24:12.829565+00:00"},{"alias_kind":"pith_short_16","alias_value":"EIKKKWLHVJESAH2V","created_at":"2026-07-05T05:24:12.829565+00:00"},{"alias_kind":"pith_short_8","alias_value":"EIKKKWLH","created_at":"2026-07-05T05:24:12.829565+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.01626","citing_title":"IMWM: Intuition Models Complement World Models for Latent Planning","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26689","citing_title":"Atomic-Probe Governance for Skill Updates in Compositional Robot Policies","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2604.26689","citing_title":"Atomic-Probe Governance for Skill Updates in Compositional Robot Policies","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24558","citing_title":"Hierarchical Behaviour Spaces","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08059","citing_title":"Governed Capability Evolution: Lifecycle-Time Compatibility Checking and Rollback for AI-Component-Based Systems, with Embodied Agents as Case Study","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08059","citing_title":"Governed Capability Evolution: Lifecycle-Time Compatibility Checking and Rollback for AI-Component-Based Systems, with Embodied Agents as Case Study","ref_index":32,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY","json":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY.json","graph_json":"https://pith.science/api/pith-number/EIKKKWLHVJESAH2VXMV7X64LUY/graph.json","events_json":"https://pith.science/api/pith-number/EIKKKWLHVJESAH2VXMV7X64LUY/events.json","paper":"https://pith.science/paper/EIKKKWLH"},"agent_actions":{"view_html":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY","download_json":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY.json","view_paper":"https://pith.science/paper/EIKKKWLH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2207.07560&json=true","fetch_graph":"https://pith.science/api/pith-number/EIKKKWLHVJESAH2VXMV7X64LUY/graph.json","fetch_events":"https://pith.science/api/pith-number/EIKKKWLHVJESAH2VXMV7X64LUY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY/action/storage_attestation","attest_author":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY/action/author_attestation","sign_citation":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY/action/citation_signature","submit_replication":"https://pith.science/pith/EIKKKWLHVJESAH2VXMV7X64LUY/action/replication_record"}},"created_at":"2026-07-05T05:24:12.829565+00:00","updated_at":"2026-07-05T05:24:12.829565+00:00"}