{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:QVPCGZBFVIW4WMBHI2SDEBO4V6","short_pith_number":"pith:QVPCGZBF","schema_version":"1.0","canonical_sha256":"855e236425aa2dcb302746a43205dcafaa9778f17dcb4d891359978043893454","source":{"kind":"arxiv","id":"2310.16029","version":1},"attestation_state":"computed","paper":{"title":"Finetuning Offline World Models in the Real World","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.RO"],"primary_cat":"cs.LG","authors_text":"Chandramouli Rajagopalan, Nicklas Hansen, Xiaolong Wang, Yunhai Feng, Ziyan Xiong","submitted_at":"2023-10-24T17:46:12Z","abstract_excerpt":"Reinforcement Learning (RL) is notoriously data-inefficient, which makes training on a real robot difficult. While model-based RL algorithms (world models) improve data-efficiency to some extent, they still require hours or days of interaction to learn skills. Recently, offline RL has been proposed as a framework for training RL policies on pre-existing datasets without any online interaction. However, constraining an algorithm to a fixed dataset induces a state-action distribution shift between training and inference, and limits its applicability to new tasks. In this work, we seek to get the"},"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":"2310.16029","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-24T17:46:12Z","cross_cats_sorted":["cs.AI","cs.CV","cs.RO"],"title_canon_sha256":"9e101ba0954487e3094c96997ea78cb77b7208c5b001dcc016486dc0969e13d9","abstract_canon_sha256":"ac87e2255521621bcf69cc46c663dca01619af196791d3196bea718bc3a8a954"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:04:34.040739Z","signature_b64":"9IgBsy5VPwx9BH08YXJCOgg0sYKwM1jIJnYlFEs7XmL0QUqgfmJ9YTghMIvuOX317givik3DQQgNFyI6NHy0CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"855e236425aa2dcb302746a43205dcafaa9778f17dcb4d891359978043893454","last_reissued_at":"2026-07-05T07:04:34.040245Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:04:34.040245Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Finetuning Offline World Models in the Real World","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.RO"],"primary_cat":"cs.LG","authors_text":"Chandramouli Rajagopalan, Nicklas Hansen, Xiaolong Wang, Yunhai Feng, Ziyan Xiong","submitted_at":"2023-10-24T17:46:12Z","abstract_excerpt":"Reinforcement Learning (RL) is notoriously data-inefficient, which makes training on a real robot difficult. While model-based RL algorithms (world models) improve data-efficiency to some extent, they still require hours or days of interaction to learn skills. Recently, offline RL has been proposed as a framework for training RL policies on pre-existing datasets without any online interaction. However, constraining an algorithm to a fixed dataset induces a state-action distribution shift between training and inference, and limits its applicability to new tasks. In this work, we seek to get the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.16029","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/2310.16029/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":"2310.16029","created_at":"2026-07-05T07:04:34.040304+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.16029v1","created_at":"2026-07-05T07:04:34.040304+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.16029","created_at":"2026-07-05T07:04:34.040304+00:00"},{"alias_kind":"pith_short_12","alias_value":"QVPCGZBFVIW4","created_at":"2026-07-05T07:04:34.040304+00:00"},{"alias_kind":"pith_short_16","alias_value":"QVPCGZBFVIW4WMBH","created_at":"2026-07-05T07:04:34.040304+00:00"},{"alias_kind":"pith_short_8","alias_value":"QVPCGZBF","created_at":"2026-07-05T07:04:34.040304+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19328","citing_title":"UBP2: Uncertainty-Balanced Preference Planning for Efficient Preference-based Reinforcement Learning","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10825","citing_title":"MODIP: Efficient Model-Based Optimization for Diffusion Policies","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2603.15759","citing_title":"Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02911","citing_title":"Learning Task-Invariant Properties via Dreamer: Enabling Efficient Policy Transfer for Quadruped Robots","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2403.09631","citing_title":"3D-VLA: A 3D Vision-Language-Action Generative World Model","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06425","citing_title":"Neural Computers","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6","json":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6.json","graph_json":"https://pith.science/api/pith-number/QVPCGZBFVIW4WMBHI2SDEBO4V6/graph.json","events_json":"https://pith.science/api/pith-number/QVPCGZBFVIW4WMBHI2SDEBO4V6/events.json","paper":"https://pith.science/paper/QVPCGZBF"},"agent_actions":{"view_html":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6","download_json":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6.json","view_paper":"https://pith.science/paper/QVPCGZBF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.16029&json=true","fetch_graph":"https://pith.science/api/pith-number/QVPCGZBFVIW4WMBHI2SDEBO4V6/graph.json","fetch_events":"https://pith.science/api/pith-number/QVPCGZBFVIW4WMBHI2SDEBO4V6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6/action/storage_attestation","attest_author":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6/action/author_attestation","sign_citation":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6/action/citation_signature","submit_replication":"https://pith.science/pith/QVPCGZBFVIW4WMBHI2SDEBO4V6/action/replication_record"}},"created_at":"2026-07-05T07:04:34.040304+00:00","updated_at":"2026-07-05T07:04:34.040304+00:00"}