{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:GSBFVVHX3G2BZDLAEZW4NHEEIP","short_pith_number":"pith:GSBFVVHX","schema_version":"1.0","canonical_sha256":"34825ad4f7d9b41c8d60266dc69c8443c44c093a0a5256a91cba8918b28e2f39","source":{"kind":"arxiv","id":"2502.20391","version":1},"attestation_state":"computed","paper":{"title":"Point Policy: Unifying Observations and Actions with Key Points for Robot Manipulation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Lerrel Pinto, Siddhant Haldar","submitted_at":"2025-02-27T18:59:18Z","abstract_excerpt":"Building robotic agents capable of operating across diverse environments and object types remains a significant challenge, often requiring extensive data collection. This is particularly restrictive in robotics, where each data point must be physically executed in the real world. Consequently, there is a critical need for alternative data sources for robotics and frameworks that enable learning from such data. In this work, we present Point Policy, a new method for learning robot policies exclusively from offline human demonstration videos and without any teleoperation data. Point Policy lever"},"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":"2502.20391","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-02-27T18:59:18Z","cross_cats_sorted":[],"title_canon_sha256":"91429b7641e39c19a760d56afb958d82e5d26512e9fcfdd6e1f65e98ee563c9f","abstract_canon_sha256":"95a5a1b42d158291f00f55fb475c148f1c8390fbeb4c03692a5431047a0d5b6f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:21:14.144363Z","signature_b64":"Ls3H5FBo9gXmUCDyZml3bnq3G72iYn3tUF8a4VCu1g1DfRK8PMJFdYHgRenjjVG2P19Wcr5xcXDKxOflYuaSDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"34825ad4f7d9b41c8d60266dc69c8443c44c093a0a5256a91cba8918b28e2f39","last_reissued_at":"2026-07-05T10:21:14.143670Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:21:14.143670Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Point Policy: Unifying Observations and Actions with Key Points for Robot Manipulation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Lerrel Pinto, Siddhant Haldar","submitted_at":"2025-02-27T18:59:18Z","abstract_excerpt":"Building robotic agents capable of operating across diverse environments and object types remains a significant challenge, often requiring extensive data collection. This is particularly restrictive in robotics, where each data point must be physically executed in the real world. Consequently, there is a critical need for alternative data sources for robotics and frameworks that enable learning from such data. In this work, we present Point Policy, a new method for learning robot policies exclusively from offline human demonstration videos and without any teleoperation data. Point Policy lever"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.20391","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/2502.20391/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":"2502.20391","created_at":"2026-07-05T10:21:14.143758+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.20391v1","created_at":"2026-07-05T10:21:14.143758+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.20391","created_at":"2026-07-05T10:21:14.143758+00:00"},{"alias_kind":"pith_short_12","alias_value":"GSBFVVHX3G2B","created_at":"2026-07-05T10:21:14.143758+00:00"},{"alias_kind":"pith_short_16","alias_value":"GSBFVVHX3G2BZDLA","created_at":"2026-07-05T10:21:14.143758+00:00"},{"alias_kind":"pith_short_8","alias_value":"GSBFVVHX","created_at":"2026-07-05T10:21:14.143758+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08436","citing_title":"EgoWAM: World Action Models Beyond Pixels with In-the-Wild Egocentric Human Data","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2606.22113","citing_title":"KITE: Decoupling Kinematics and Interaction for Zero-Shot Cross-Embodiment Manipulation","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17054","citing_title":"Human Universal Grasping","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13515","citing_title":"MaskWAM: Unifying Mask Prompting and Prediction for World-Action Models","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11628","citing_title":"LUCID: Learning Embodiment-Agnostic Intent Models from Unstructured Human Videos for Scalable Dexterous Robot Skill Acquisition","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10743","citing_title":"Hand-centric Human-to-Robot Trajectory Transfer from Video Demonstrations via Open-World Contact Localization","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10025","citing_title":"GHOST: Hierarchical Sub-Goal Policies for Generalizing Robot Manipulation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09314","citing_title":"KPGrasp: Scalable Keypoint Flow Matching for Dexterous Grasp Generation","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12369","citing_title":"GuidedVLA: Specifying Task-Relevant Factors via Plug-and-Play Action Attention Specialization","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26649","citing_title":"On the Generalization Capabilities, Design Choices and Limitations of Keypoint Imitation Learning","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16743","citing_title":"LACE: Latent Visual Representation for Cross-Embodiment Learning","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2511.04671","citing_title":"X-Diffusion: Training Diffusion Policies on Cross-Embodiment Human Demonstrations","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12369","citing_title":"GuidedVLA: Specifying Task-Relevant Factors via Plug-and-Play Action Attention Specialization","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10809","citing_title":"WARPED: Wrist-Aligned Rendering for Robot Policy Learning from Egocentric Human Demonstrations","ref_index":34,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP","json":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP.json","graph_json":"https://pith.science/api/pith-number/GSBFVVHX3G2BZDLAEZW4NHEEIP/graph.json","events_json":"https://pith.science/api/pith-number/GSBFVVHX3G2BZDLAEZW4NHEEIP/events.json","paper":"https://pith.science/paper/GSBFVVHX"},"agent_actions":{"view_html":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP","download_json":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP.json","view_paper":"https://pith.science/paper/GSBFVVHX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.20391&json=true","fetch_graph":"https://pith.science/api/pith-number/GSBFVVHX3G2BZDLAEZW4NHEEIP/graph.json","fetch_events":"https://pith.science/api/pith-number/GSBFVVHX3G2BZDLAEZW4NHEEIP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP/action/storage_attestation","attest_author":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP/action/author_attestation","sign_citation":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP/action/citation_signature","submit_replication":"https://pith.science/pith/GSBFVVHX3G2BZDLAEZW4NHEEIP/action/replication_record"}},"created_at":"2026-07-05T10:21:14.143758+00:00","updated_at":"2026-07-05T10:21:14.143758+00:00"}