{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:SPDPFBJWCSALPULZOVB4AFVLDV","short_pith_number":"pith:SPDPFBJW","schema_version":"1.0","canonical_sha256":"93c6f285361480b7d1797543c016ab1d5b48fe8f8875752b8fb9fcb21af4a5de","source":{"kind":"arxiv","id":"2210.03109","version":1},"attestation_state":"computed","paper":{"title":"Real-World Robot Learning with Masked Visual Pre-training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.RO","authors_text":"Ilija Radosavovic, Jitendra Malik, Pieter Abbeel, Stephen James, Tete Xiao, Trevor Darrell","submitted_at":"2022-10-06T17:59:01Z","abstract_excerpt":"In this work, we explore self-supervised visual pre-training on images from diverse, in-the-wild videos for real-world robotic tasks. Like prior work, our visual representations are pre-trained via a masked autoencoder (MAE), frozen, and then passed into a learnable control module. Unlike prior work, we show that the pre-trained representations are effective across a range of real-world robotic tasks and embodiments. We find that our encoder consistently outperforms CLIP (up to 75%), supervised ImageNet pre-training (up to 81%), and training from scratch (up to 81%). Finally, we train a 307M p"},"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":"2210.03109","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2022-10-06T17:59:01Z","cross_cats_sorted":["cs.CV","cs.LG"],"title_canon_sha256":"4d639d3a5892d45ffba995a55d1a01ea336279648c3b3029ccd98d3522c161db","abstract_canon_sha256":"cfd650cc8abfa4fb901aa1cdbbe15e1b56f9dd2c3e40f0c761a41679604e30da"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:04:07.711307Z","signature_b64":"oaWToVEBKIPjHISbRACDQFtzPBpLPAKYvTBixDid6jrJh57XiBFLKLyuZTgg15ydab32cu+FQpuLG6ip46MSCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"93c6f285361480b7d1797543c016ab1d5b48fe8f8875752b8fb9fcb21af4a5de","last_reissued_at":"2026-07-05T05:04:07.710834Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:04:07.710834Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Real-World Robot Learning with Masked Visual Pre-training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.RO","authors_text":"Ilija Radosavovic, Jitendra Malik, Pieter Abbeel, Stephen James, Tete Xiao, Trevor Darrell","submitted_at":"2022-10-06T17:59:01Z","abstract_excerpt":"In this work, we explore self-supervised visual pre-training on images from diverse, in-the-wild videos for real-world robotic tasks. Like prior work, our visual representations are pre-trained via a masked autoencoder (MAE), frozen, and then passed into a learnable control module. Unlike prior work, we show that the pre-trained representations are effective across a range of real-world robotic tasks and embodiments. We find that our encoder consistently outperforms CLIP (up to 75%), supervised ImageNet pre-training (up to 81%), and training from scratch (up to 81%). Finally, we train a 307M p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2210.03109","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/2210.03109/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":"2210.03109","created_at":"2026-07-05T05:04:07.710893+00:00"},{"alias_kind":"arxiv_version","alias_value":"2210.03109v1","created_at":"2026-07-05T05:04:07.710893+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2210.03109","created_at":"2026-07-05T05:04:07.710893+00:00"},{"alias_kind":"pith_short_12","alias_value":"SPDPFBJWCSAL","created_at":"2026-07-05T05:04:07.710893+00:00"},{"alias_kind":"pith_short_16","alias_value":"SPDPFBJWCSALPULZ","created_at":"2026-07-05T05:04:07.710893+00:00"},{"alias_kind":"pith_short_8","alias_value":"SPDPFBJW","created_at":"2026-07-05T05:04:07.710893+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.17256","citing_title":"Contrastive Action-Image Pre-training for Visuomotor Control","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2511.14427","citing_title":"Self-Supervised Multisensory Pretraining for Contact-Rich Robot Reinforcement Learning","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV","json":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV.json","graph_json":"https://pith.science/api/pith-number/SPDPFBJWCSALPULZOVB4AFVLDV/graph.json","events_json":"https://pith.science/api/pith-number/SPDPFBJWCSALPULZOVB4AFVLDV/events.json","paper":"https://pith.science/paper/SPDPFBJW"},"agent_actions":{"view_html":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV","download_json":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV.json","view_paper":"https://pith.science/paper/SPDPFBJW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2210.03109&json=true","fetch_graph":"https://pith.science/api/pith-number/SPDPFBJWCSALPULZOVB4AFVLDV/graph.json","fetch_events":"https://pith.science/api/pith-number/SPDPFBJWCSALPULZOVB4AFVLDV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV/action/storage_attestation","attest_author":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV/action/author_attestation","sign_citation":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV/action/citation_signature","submit_replication":"https://pith.science/pith/SPDPFBJWCSALPULZOVB4AFVLDV/action/replication_record"}},"created_at":"2026-07-05T05:04:07.710893+00:00","updated_at":"2026-07-05T05:04:07.710893+00:00"}