{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6ROE7RLFY5DSYHWYKF3EHMRXAK","short_pith_number":"pith:6ROE7RLF","schema_version":"1.0","canonical_sha256":"f45c4fc565c7472c1ed8517643b2370283b45524daa1de28ecbf6481c78dc3f4","source":{"kind":"arxiv","id":"2405.03685","version":1},"attestation_state":"computed","paper":{"title":"Language-Image Models with 3D Understanding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Boris Ivanovic, Boyi Li, Edward Schmerling, Jang Hyun Cho, Marco Pavone, Philipp Kr\\\"ahenb\\\"uhl, Xinshuo Weng, Yan Wang, Yue Wang, Yulong Cao, Yurong You","submitted_at":"2024-05-06T17:57:27Z","abstract_excerpt":"Multi-modal large language models (MLLMs) have shown incredible capabilities in a variety of 2D vision and language tasks. We extend MLLMs' perceptual capabilities to ground and reason about images in 3-dimensional space. To that end, we first develop a large-scale pre-training dataset for 2D and 3D called LV3D by combining multiple existing 2D and 3D recognition datasets under a common task formulation: as multi-turn question-answering. Next, we introduce a new MLLM named Cube-LLM and pre-train it on LV3D. We show that pure data scaling makes a strong 3D perception capability without 3D speci"},"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":"2405.03685","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-06T17:57:27Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"49e3e559649ef7174b47f91b852c1d799185898c269cdb0a9b411bde3de646ef","abstract_canon_sha256":"95c57e54fca5754747449c268cca74aa0a8e08198c37601605caab831ad105b5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:16:06.385425Z","signature_b64":"NMoPYkmgGcXWtcPffDS+M/DWFN3wZOwBnRllNkv4kVEdir2gYS65J+mIBLGfK2C69ZYQLlQrbqzuM7KkjcfFDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f45c4fc565c7472c1ed8517643b2370283b45524daa1de28ecbf6481c78dc3f4","last_reissued_at":"2026-07-05T08:16:06.384937Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:16:06.384937Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Language-Image Models with 3D Understanding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Boris Ivanovic, Boyi Li, Edward Schmerling, Jang Hyun Cho, Marco Pavone, Philipp Kr\\\"ahenb\\\"uhl, Xinshuo Weng, Yan Wang, Yue Wang, Yulong Cao, Yurong You","submitted_at":"2024-05-06T17:57:27Z","abstract_excerpt":"Multi-modal large language models (MLLMs) have shown incredible capabilities in a variety of 2D vision and language tasks. We extend MLLMs' perceptual capabilities to ground and reason about images in 3-dimensional space. To that end, we first develop a large-scale pre-training dataset for 2D and 3D called LV3D by combining multiple existing 2D and 3D recognition datasets under a common task formulation: as multi-turn question-answering. Next, we introduce a new MLLM named Cube-LLM and pre-train it on LV3D. We show that pure data scaling makes a strong 3D perception capability without 3D speci"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.03685","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/2405.03685/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":"2405.03685","created_at":"2026-07-05T08:16:06.384991+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.03685v1","created_at":"2026-07-05T08:16:06.384991+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.03685","created_at":"2026-07-05T08:16:06.384991+00:00"},{"alias_kind":"pith_short_12","alias_value":"6ROE7RLFY5DS","created_at":"2026-07-05T08:16:06.384991+00:00"},{"alias_kind":"pith_short_16","alias_value":"6ROE7RLFY5DSYHWY","created_at":"2026-07-05T08:16:06.384991+00:00"},{"alias_kind":"pith_short_8","alias_value":"6ROE7RLF","created_at":"2026-07-05T08:16:06.384991+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.16713","citing_title":"GeoWorld-VLM: Geometry from World Models for Vision-Language Models","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2602.19710","citing_title":"PoseVLA: Universal Pose Pretraining for Generalizable Vision-Language-Action Policies","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16713","citing_title":"GeoWorld-VLM: Geometry from World Models for Vision-Language Models","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2511.10946","citing_title":"Abstract 3D Perception for Spatial Intelligence in Vision-Language Models","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2410.23262","citing_title":"EMMA: End-to-End Multimodal Model for Autonomous Driving","ref_index":197,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05212","citing_title":"Boxer: Robust Lifting of Open-World 2D Bounding Boxes to 3D","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK","json":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK.json","graph_json":"https://pith.science/api/pith-number/6ROE7RLFY5DSYHWYKF3EHMRXAK/graph.json","events_json":"https://pith.science/api/pith-number/6ROE7RLFY5DSYHWYKF3EHMRXAK/events.json","paper":"https://pith.science/paper/6ROE7RLF"},"agent_actions":{"view_html":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK","download_json":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK.json","view_paper":"https://pith.science/paper/6ROE7RLF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.03685&json=true","fetch_graph":"https://pith.science/api/pith-number/6ROE7RLFY5DSYHWYKF3EHMRXAK/graph.json","fetch_events":"https://pith.science/api/pith-number/6ROE7RLFY5DSYHWYKF3EHMRXAK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK/action/storage_attestation","attest_author":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK/action/author_attestation","sign_citation":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK/action/citation_signature","submit_replication":"https://pith.science/pith/6ROE7RLFY5DSYHWYKF3EHMRXAK/action/replication_record"}},"created_at":"2026-07-05T08:16:06.384991+00:00","updated_at":"2026-07-05T08:16:06.384991+00:00"}