{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SHUC4EYGNGABJRA5LXRERXLIY7","short_pith_number":"pith:SHUC4EYG","schema_version":"1.0","canonical_sha256":"91e82e1306698014c41d5de248dd68c7c60b6cd6f89a0ea55dd7c7679ac4340c","source":{"kind":"arxiv","id":"2503.08593","version":1},"attestation_state":"computed","paper":{"title":"Proc4Gem: Foundation models for physical agency through procedural generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Adil Dostmohamed, Adrian Collister, Arunkumar Byravan, Ben Moran, Catarina Barros, Daniel Zheng, Francesco Nori, Francesco Romano, Jan Humplik, Jose Enrique Chen, Joss Moore, Ken Caluwaerts, Kieran Connell, Leonard Hasenclever, Marissa Giustina, Matt Young, Nicolas Heess, Sandy H. Huang, Stefano Saliceti, Steven Bohez, Yixin Lin","submitted_at":"2025-03-11T16:29:44Z","abstract_excerpt":"In robot learning, it is common to either ignore the environment semantics, focusing on tasks like whole-body control which only require reasoning about robot-environment contacts, or conversely to ignore contact dynamics, focusing on grounding high-level movement in vision and language. In this work, we show that advances in generative modeling, photorealistic rendering, and procedural generation allow us to tackle tasks requiring both. By generating contact-rich trajectories with accurate physics in semantically-diverse simulations, we can distill behaviors into large multimodal models that "},"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":"2503.08593","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-03-11T16:29:44Z","cross_cats_sorted":[],"title_canon_sha256":"242f9cf40dd3c2e3ff8c8835ff9066dd70fac3ca1302b985d88eb393bde959c6","abstract_canon_sha256":"70f75edfd546d5889ae3b43d8841141858ebb16437671933e4ac1a848036fd88"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:29:03.195760Z","signature_b64":"DlLyP4bQ1XIlSAh3ITWSltkadDbhe0AZnTEEagNaNc/PATO57YOSZSB4Qc+iFoL4DpxKN5W3Va4C0AfoRU8HBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"91e82e1306698014c41d5de248dd68c7c60b6cd6f89a0ea55dd7c7679ac4340c","last_reissued_at":"2026-07-05T10:29:03.195225Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:29:03.195225Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Proc4Gem: Foundation models for physical agency through procedural generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Adil Dostmohamed, Adrian Collister, Arunkumar Byravan, Ben Moran, Catarina Barros, Daniel Zheng, Francesco Nori, Francesco Romano, Jan Humplik, Jose Enrique Chen, Joss Moore, Ken Caluwaerts, Kieran Connell, Leonard Hasenclever, Marissa Giustina, Matt Young, Nicolas Heess, Sandy H. Huang, Stefano Saliceti, Steven Bohez, Yixin Lin","submitted_at":"2025-03-11T16:29:44Z","abstract_excerpt":"In robot learning, it is common to either ignore the environment semantics, focusing on tasks like whole-body control which only require reasoning about robot-environment contacts, or conversely to ignore contact dynamics, focusing on grounding high-level movement in vision and language. In this work, we show that advances in generative modeling, photorealistic rendering, and procedural generation allow us to tackle tasks requiring both. By generating contact-rich trajectories with accurate physics in semantically-diverse simulations, we can distill behaviors into large multimodal models that "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.08593","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/2503.08593/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":"2503.08593","created_at":"2026-07-05T10:29:03.195291+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.08593v1","created_at":"2026-07-05T10:29:03.195291+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.08593","created_at":"2026-07-05T10:29:03.195291+00:00"},{"alias_kind":"pith_short_12","alias_value":"SHUC4EYGNGAB","created_at":"2026-07-05T10:29:03.195291+00:00"},{"alias_kind":"pith_short_16","alias_value":"SHUC4EYGNGABJRA5","created_at":"2026-07-05T10:29:03.195291+00:00"},{"alias_kind":"pith_short_8","alias_value":"SHUC4EYG","created_at":"2026-07-05T10:29:03.195291+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.10087","citing_title":"Foundation Model Driven Robotics: A Comprehensive Review","ref_index":126,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7","json":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7.json","graph_json":"https://pith.science/api/pith-number/SHUC4EYGNGABJRA5LXRERXLIY7/graph.json","events_json":"https://pith.science/api/pith-number/SHUC4EYGNGABJRA5LXRERXLIY7/events.json","paper":"https://pith.science/paper/SHUC4EYG"},"agent_actions":{"view_html":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7","download_json":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7.json","view_paper":"https://pith.science/paper/SHUC4EYG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.08593&json=true","fetch_graph":"https://pith.science/api/pith-number/SHUC4EYGNGABJRA5LXRERXLIY7/graph.json","fetch_events":"https://pith.science/api/pith-number/SHUC4EYGNGABJRA5LXRERXLIY7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7/action/storage_attestation","attest_author":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7/action/author_attestation","sign_citation":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7/action/citation_signature","submit_replication":"https://pith.science/pith/SHUC4EYGNGABJRA5LXRERXLIY7/action/replication_record"}},"created_at":"2026-07-05T10:29:03.195291+00:00","updated_at":"2026-07-05T10:29:03.195291+00:00"}