{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PFUQQ4JETGZE2EYF6CWT5QD7DM","short_pith_number":"pith:PFUQQ4JE","schema_version":"1.0","canonical_sha256":"796908712499b24d1305f0ad3ec07f1b27010de045c2850f0d9f022252cafd5f","source":{"kind":"arxiv","id":"2503.19757","version":2},"attestation_state":"computed","paper":{"title":"Dita: Scaling Diffusion Transformer for Generalist Vision-Language-Action Policy","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.RO","authors_text":"Chengyang Zhao, Haonan Duan, Hengjun Pu, Jifeng Dai, Ronglei Tong, Tianyi Zhang, Xizhou Zhu, Yuntao Chen, Yu Qiao, Yuwen Xiong, Zhi Hou","submitted_at":"2025-03-25T15:19:56Z","abstract_excerpt":"While recent vision-language-action models trained on diverse robot datasets exhibit promising generalization capabilities with limited in-domain data, their reliance on compact action heads to predict discretized or continuous actions constrains adaptability to heterogeneous action spaces. We present Dita, a scalable framework that leverages Transformer architectures to directly denoise continuous action sequences through a unified multimodal diffusion process. Departing from prior methods that condition denoising on fused embeddings via shallow networks, Dita employs in-context conditioning "},"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":"2503.19757","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2025-03-25T15:19:56Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"4c1b39aebdda1a75a8bfd8658fd6cde20937c267d8ce434aacff2ead47dd8da3","abstract_canon_sha256":"9565dc7af7367b2a63e02ad1cf89a2249b9d5b7c87e6865f7065d2200ca98d52"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:06:08.333874Z","signature_b64":"oAQejFZ4JxrNv6kkT4y4/VB2TW/w2iiK3WR8TCOls5JGE/GqgrLFaX/GXpJ0b1V5zSX+ELNNeohuoIO5o6bRBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"796908712499b24d1305f0ad3ec07f1b27010de045c2850f0d9f022252cafd5f","last_reissued_at":"2026-07-05T12:06:08.333354Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:06:08.333354Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dita: Scaling Diffusion Transformer for Generalist Vision-Language-Action Policy","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.RO","authors_text":"Chengyang Zhao, Haonan Duan, Hengjun Pu, Jifeng Dai, Ronglei Tong, Tianyi Zhang, Xizhou Zhu, Yuntao Chen, Yu Qiao, Yuwen Xiong, Zhi Hou","submitted_at":"2025-03-25T15:19:56Z","abstract_excerpt":"While recent vision-language-action models trained on diverse robot datasets exhibit promising generalization capabilities with limited in-domain data, their reliance on compact action heads to predict discretized or continuous actions constrains adaptability to heterogeneous action spaces. We present Dita, a scalable framework that leverages Transformer architectures to directly denoise continuous action sequences through a unified multimodal diffusion process. Departing from prior methods that condition denoising on fused embeddings via shallow networks, Dita employs in-context conditioning "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.19757","kind":"arxiv","version":2},"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/2503.19757/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":"2503.19757","created_at":"2026-07-05T12:06:08.333413+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.19757v2","created_at":"2026-07-05T12:06:08.333413+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.19757","created_at":"2026-07-05T12:06:08.333413+00:00"},{"alias_kind":"pith_short_12","alias_value":"PFUQQ4JETGZE","created_at":"2026-07-05T12:06:08.333413+00:00"},{"alias_kind":"pith_short_16","alias_value":"PFUQQ4JETGZE2EYF","created_at":"2026-07-05T12:06:08.333413+00:00"},{"alias_kind":"pith_short_8","alias_value":"PFUQQ4JE","created_at":"2026-07-05T12:06:08.333413+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2505.17016","citing_title":"Interactive Post-Training for Vision-Language-Action Models","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17486","citing_title":"DyGRO-VLA: Cross-Task Scaling of Vision-Language-Action Models via Dynamic Grouped Residual Optimization","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2506.13456","citing_title":"Block-wise Adaptive Caching for Accelerating Diffusion Policy","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2511.14148","citing_title":"AsyncVLA: Asynchronous Flow Matching for Vision-Language-Action Models","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2507.04447","citing_title":"DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledge","ref_index":91,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11665","citing_title":"Nautilus: From One Prompt to Plug-and-Play Robot Learning","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2601.16163","citing_title":"Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05126","citing_title":"ConsisVLA-4D: Advancing Spatiotemporal Consistency in Efficient 3D-Perception and 4D-Reasoning for Robotic Manipulation","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21331","citing_title":"FingerViP: Learning Real-World Dexterous Manipulation with Fingertip Visual Perception","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07381","citing_title":"Escaping the Diversity Trap in Robotic Manipulation via Anchor-Centric Adaptation","ref_index":59,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM","json":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM.json","graph_json":"https://pith.science/api/pith-number/PFUQQ4JETGZE2EYF6CWT5QD7DM/graph.json","events_json":"https://pith.science/api/pith-number/PFUQQ4JETGZE2EYF6CWT5QD7DM/events.json","paper":"https://pith.science/paper/PFUQQ4JE"},"agent_actions":{"view_html":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM","download_json":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM.json","view_paper":"https://pith.science/paper/PFUQQ4JE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.19757&json=true","fetch_graph":"https://pith.science/api/pith-number/PFUQQ4JETGZE2EYF6CWT5QD7DM/graph.json","fetch_events":"https://pith.science/api/pith-number/PFUQQ4JETGZE2EYF6CWT5QD7DM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM/action/storage_attestation","attest_author":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM/action/author_attestation","sign_citation":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM/action/citation_signature","submit_replication":"https://pith.science/pith/PFUQQ4JETGZE2EYF6CWT5QD7DM/action/replication_record"}},"created_at":"2026-07-05T12:06:08.333413+00:00","updated_at":"2026-07-05T12:06:08.333413+00:00"}