{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:L35SFJBEVJFPPPRK4KKPPERCKD","short_pith_number":"pith:L35SFJBE","schema_version":"1.0","canonical_sha256":"5efb22a424aa4af7be2ae294f7922250f97e46d554e3689fe10c4f7afc6edc76","source":{"kind":"arxiv","id":"2506.18890","version":1},"attestation_state":"computed","paper":{"title":"4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chen Ziwen, Hao Tan, Jianing Yang, Joyce Chai, Kai Zhang, Kalyan Sunkavalli, Mohit Bansal, Sai Bi, Shoubin Yu, Sihan Xu, Xuweiyi Chen, Zexiang Xu, Ziqiao Ma","submitted_at":"2025-06-23T17:57:47Z","abstract_excerpt":"Can we scale 4D pretraining to learn general space-time representations that reconstruct an object from a few views at some times to any view at any time? We provide an affirmative answer with 4D-LRM, the first large-scale 4D reconstruction model that takes input from unconstrained views and timestamps and renders arbitrary novel view-time combinations. Unlike prior 4D approaches, e.g., optimization-based, geometry-based, or generative, that struggle with efficiency, generalization, or faithfulness, 4D-LRM learns a unified space-time representation and directly predicts per-pixel 4D Gaussian 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":"2506.18890","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-23T17:57:47Z","cross_cats_sorted":[],"title_canon_sha256":"97548918ab4782e81a6bd7f9a09748ef085aaa6343610c4ae1f7e4fac1c0f54a","abstract_canon_sha256":"c5b901aa477dffce025d011c57659fc6f53fb095d7eff1e23929ceaa33d1a28e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:25:57.663277Z","signature_b64":"vN/k5bvV/XuLSbCCfLtCbJDw8vxtHBb7RvJWM9auDDwj2rDZGghwNYChSEsP1ep5nEiUA5ikSfhEISfEcxQNDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5efb22a424aa4af7be2ae294f7922250f97e46d554e3689fe10c4f7afc6edc76","last_reissued_at":"2026-07-05T11:25:57.662811Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:25:57.662811Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chen Ziwen, Hao Tan, Jianing Yang, Joyce Chai, Kai Zhang, Kalyan Sunkavalli, Mohit Bansal, Sai Bi, Shoubin Yu, Sihan Xu, Xuweiyi Chen, Zexiang Xu, Ziqiao Ma","submitted_at":"2025-06-23T17:57:47Z","abstract_excerpt":"Can we scale 4D pretraining to learn general space-time representations that reconstruct an object from a few views at some times to any view at any time? We provide an affirmative answer with 4D-LRM, the first large-scale 4D reconstruction model that takes input from unconstrained views and timestamps and renders arbitrary novel view-time combinations. Unlike prior 4D approaches, e.g., optimization-based, geometry-based, or generative, that struggle with efficiency, generalization, or faithfulness, 4D-LRM learns a unified space-time representation and directly predicts per-pixel 4D Gaussian p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.18890","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/2506.18890/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":"2506.18890","created_at":"2026-07-05T11:25:57.662870+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.18890v1","created_at":"2026-07-05T11:25:57.662870+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.18890","created_at":"2026-07-05T11:25:57.662870+00:00"},{"alias_kind":"pith_short_12","alias_value":"L35SFJBEVJFP","created_at":"2026-07-05T11:25:57.662870+00:00"},{"alias_kind":"pith_short_16","alias_value":"L35SFJBEVJFPPPRK","created_at":"2026-07-05T11:25:57.662870+00:00"},{"alias_kind":"pith_short_8","alias_value":"L35SFJBE","created_at":"2026-07-05T11:25:57.662870+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.28215","citing_title":"HAT-4D: Lifting Monocular Video for 4D Multi-Object Interactions via Human-Agent Collaboration","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31595","citing_title":"Learning Global Motion with Compact Gaussians for Feed-Forward 4D Reconstruction","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05182","citing_title":"LSRM: High-Fidelity Object-Centric Reconstruction via Scaled Context Windows","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14025","citing_title":"Feed-Forward 3D Scene Modeling: A Problem-Driven Perspective","ref_index":190,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD","json":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD.json","graph_json":"https://pith.science/api/pith-number/L35SFJBEVJFPPPRK4KKPPERCKD/graph.json","events_json":"https://pith.science/api/pith-number/L35SFJBEVJFPPPRK4KKPPERCKD/events.json","paper":"https://pith.science/paper/L35SFJBE"},"agent_actions":{"view_html":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD","download_json":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD.json","view_paper":"https://pith.science/paper/L35SFJBE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.18890&json=true","fetch_graph":"https://pith.science/api/pith-number/L35SFJBEVJFPPPRK4KKPPERCKD/graph.json","fetch_events":"https://pith.science/api/pith-number/L35SFJBEVJFPPPRK4KKPPERCKD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD/action/storage_attestation","attest_author":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD/action/author_attestation","sign_citation":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD/action/citation_signature","submit_replication":"https://pith.science/pith/L35SFJBEVJFPPPRK4KKPPERCKD/action/replication_record"}},"created_at":"2026-07-05T11:25:57.662870+00:00","updated_at":"2026-07-05T11:25:57.662870+00:00"}