{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6EQTBKAKX7DFM6IVTB3K7U6KOS","short_pith_number":"pith:6EQTBKAK","schema_version":"1.0","canonical_sha256":"f12130a80abfc65679159876afd3ca74a455a966a572cb93810ce3f53f1f8bf3","source":{"kind":"arxiv","id":"2409.12192","version":2},"attestation_state":"computed","paper":{"title":"DynaMo: In-Domain Dynamics Pretraining for Visuo-Motor Control","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.LG"],"primary_cat":"cs.RO","authors_text":"Aadhithya Iyer, Hengkai Pan, Lerrel Pinto, Siddhant Haldar, Zichen Jeff Cui","submitted_at":"2024-09-18T17:59:43Z","abstract_excerpt":"Imitation learning has proven to be a powerful tool for training complex visuomotor policies. However, current methods often require hundreds to thousands of expert demonstrations to handle high-dimensional visual observations. A key reason for this poor data efficiency is that visual representations are predominantly either pretrained on out-of-domain data or trained directly through a behavior cloning objective. In this work, we present DynaMo, a new in-domain, self-supervised method for learning visual representations. Given a set of expert demonstrations, we jointly learn a latent inverse "},"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":"2409.12192","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2024-09-18T17:59:43Z","cross_cats_sorted":["cs.AI","cs.CV","cs.LG"],"title_canon_sha256":"8af22c36d22a0d1f49db8afb16f94149abd9d74c67e161d855bd8b65ee9fda5e","abstract_canon_sha256":"9d534e887b16e7628b44b558becc585829ddf0ca0732dd5e8e36ac8eb70b2a5b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:28:55.023398Z","signature_b64":"OCjGjvOFlkQ7lqSXnMMNtvxIhBp5zr7pt8X9Z5t7PId3KrhRzAkJTKRSDBIGCoQFe7PgajABfJVeHsS80FWtAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f12130a80abfc65679159876afd3ca74a455a966a572cb93810ce3f53f1f8bf3","last_reissued_at":"2026-07-05T09:28:55.022907Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:28:55.022907Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DynaMo: In-Domain Dynamics Pretraining for Visuo-Motor Control","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.LG"],"primary_cat":"cs.RO","authors_text":"Aadhithya Iyer, Hengkai Pan, Lerrel Pinto, Siddhant Haldar, Zichen Jeff Cui","submitted_at":"2024-09-18T17:59:43Z","abstract_excerpt":"Imitation learning has proven to be a powerful tool for training complex visuomotor policies. However, current methods often require hundreds to thousands of expert demonstrations to handle high-dimensional visual observations. A key reason for this poor data efficiency is that visual representations are predominantly either pretrained on out-of-domain data or trained directly through a behavior cloning objective. In this work, we present DynaMo, a new in-domain, self-supervised method for learning visual representations. Given a set of expert demonstrations, we jointly learn a latent inverse "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.12192","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/2409.12192/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":"2409.12192","created_at":"2026-07-05T09:28:55.022958+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.12192v2","created_at":"2026-07-05T09:28:55.022958+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.12192","created_at":"2026-07-05T09:28:55.022958+00:00"},{"alias_kind":"pith_short_12","alias_value":"6EQTBKAKX7DF","created_at":"2026-07-05T09:28:55.022958+00:00"},{"alias_kind":"pith_short_16","alias_value":"6EQTBKAKX7DFM6IV","created_at":"2026-07-05T09:28:55.022958+00:00"},{"alias_kind":"pith_short_8","alias_value":"6EQTBKAK","created_at":"2026-07-05T09:28:55.022958+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.06100","citing_title":"HyperVis: Continuous Latent Visual Relational Graphs on the Lorentz Hyperboloid for Compositional Reasoning","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.04130","citing_title":"CLAW: Learning Continuous Latent Action World Models via Adversarial Latent Regularization","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20223","citing_title":"Why Latent Actions Fail, and How to Prevent It","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2602.20231","citing_title":"UniLACT: Depth-Aware RGB Latent Action Learning for Vision-Language-Action Models","ref_index":8,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS","json":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS.json","graph_json":"https://pith.science/api/pith-number/6EQTBKAKX7DFM6IVTB3K7U6KOS/graph.json","events_json":"https://pith.science/api/pith-number/6EQTBKAKX7DFM6IVTB3K7U6KOS/events.json","paper":"https://pith.science/paper/6EQTBKAK"},"agent_actions":{"view_html":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS","download_json":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS.json","view_paper":"https://pith.science/paper/6EQTBKAK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.12192&json=true","fetch_graph":"https://pith.science/api/pith-number/6EQTBKAKX7DFM6IVTB3K7U6KOS/graph.json","fetch_events":"https://pith.science/api/pith-number/6EQTBKAKX7DFM6IVTB3K7U6KOS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS/action/storage_attestation","attest_author":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS/action/author_attestation","sign_citation":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS/action/citation_signature","submit_replication":"https://pith.science/pith/6EQTBKAKX7DFM6IVTB3K7U6KOS/action/replication_record"}},"created_at":"2026-07-05T09:28:55.022958+00:00","updated_at":"2026-07-05T09:28:55.022958+00:00"}