{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3A2OQLJWTGFKBGRP2LLY2T7RXX","short_pith_number":"pith:3A2OQLJW","schema_version":"1.0","canonical_sha256":"d834e82d36998aa09a2fd2d78d4ff1bdfc8cb85c3231cbf6ef52ab56c932f52a","source":{"kind":"arxiv","id":"2506.23126","version":4},"attestation_state":"computed","paper":{"title":"ParticleFormer: A 3D Point Cloud World Model for Multi-Object, Multi-Material Robotic Manipulation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Jiankai Sun, Mac Schwager, Qianzhong Chen, Suning Huang, Xiaohan Zhang","submitted_at":"2025-06-29T07:23:56Z","abstract_excerpt":"3D world models (i.e., learning-based 3D dynamics models) offer a promising approach to generalizable robotic manipulation by capturing the underlying physics of environment evolution conditioned on robot actions. However, existing 3D world models are primarily limited to single-material dynamics using a particle-based Graph Neural Network model, and often require time-consuming 3D scene reconstruction to obtain 3D particle tracks for training. In this work, we present ParticleFormer, a Transformer-based point cloud world model trained with a hybrid point cloud reconstruction loss, supervising"},"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":"2506.23126","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-06-29T07:23:56Z","cross_cats_sorted":[],"title_canon_sha256":"21a376320d12926d32f5629af552027dc5151d7e76ed4f4e56ffaae399353224","abstract_canon_sha256":"f1a9905d53f1f4bd11bb4f2305fceb74f26cdcc2bd4f67035a6ea6e6f56bf167"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:59:16.644426Z","signature_b64":"ITVpw025I6NR/9rcR4IyLHxHX4GrC4YDWbQ0RH/8OPqUUVxcYLi9MBydOzjxOCDmSmZQumMgNa4nOC3lf+9VDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d834e82d36998aa09a2fd2d78d4ff1bdfc8cb85c3231cbf6ef52ab56c932f52a","last_reissued_at":"2026-07-05T11:59:16.643942Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:59:16.643942Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ParticleFormer: A 3D Point Cloud World Model for Multi-Object, Multi-Material Robotic Manipulation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Jiankai Sun, Mac Schwager, Qianzhong Chen, Suning Huang, Xiaohan Zhang","submitted_at":"2025-06-29T07:23:56Z","abstract_excerpt":"3D world models (i.e., learning-based 3D dynamics models) offer a promising approach to generalizable robotic manipulation by capturing the underlying physics of environment evolution conditioned on robot actions. However, existing 3D world models are primarily limited to single-material dynamics using a particle-based Graph Neural Network model, and often require time-consuming 3D scene reconstruction to obtain 3D particle tracks for training. In this work, we present ParticleFormer, a Transformer-based point cloud world model trained with a hybrid point cloud reconstruction loss, supervising"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.23126","kind":"arxiv","version":4},"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/2506.23126/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":"2506.23126","created_at":"2026-07-05T11:59:16.644006+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.23126v4","created_at":"2026-07-05T11:59:16.644006+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.23126","created_at":"2026-07-05T11:59:16.644006+00:00"},{"alias_kind":"pith_short_12","alias_value":"3A2OQLJWTGFK","created_at":"2026-07-05T11:59:16.644006+00:00"},{"alias_kind":"pith_short_16","alias_value":"3A2OQLJWTGFKBGRP","created_at":"2026-07-05T11:59:16.644006+00:00"},{"alias_kind":"pith_short_8","alias_value":"3A2OQLJW","created_at":"2026-07-05T11:59:16.644006+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06925","citing_title":"Grounding Spatial Relations in a Compact World Model: Instruction Leakage and a Goal-Free Dynamics Fix","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2604.22748","citing_title":"Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond","ref_index":149,"is_internal_anchor":false},{"citing_arxiv_id":"2510.10125","citing_title":"Ctrl-World: A Controllable Generative World Model for Robot Manipulation","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23121","citing_title":"Breaking Lock-In: Preserving Steerability under Low-Data VLA Post-Training","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2604.22748","citing_title":"Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond","ref_index":149,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07079","citing_title":"Learning Visual Feature-Based World Models via Residual Latent Action","ref_index":29,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX","json":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX.json","graph_json":"https://pith.science/api/pith-number/3A2OQLJWTGFKBGRP2LLY2T7RXX/graph.json","events_json":"https://pith.science/api/pith-number/3A2OQLJWTGFKBGRP2LLY2T7RXX/events.json","paper":"https://pith.science/paper/3A2OQLJW"},"agent_actions":{"view_html":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX","download_json":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX.json","view_paper":"https://pith.science/paper/3A2OQLJW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.23126&json=true","fetch_graph":"https://pith.science/api/pith-number/3A2OQLJWTGFKBGRP2LLY2T7RXX/graph.json","fetch_events":"https://pith.science/api/pith-number/3A2OQLJWTGFKBGRP2LLY2T7RXX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX/action/storage_attestation","attest_author":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX/action/author_attestation","sign_citation":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX/action/citation_signature","submit_replication":"https://pith.science/pith/3A2OQLJWTGFKBGRP2LLY2T7RXX/action/replication_record"}},"created_at":"2026-07-05T11:59:16.644006+00:00","updated_at":"2026-07-05T11:59:16.644006+00:00"}