{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LO5BPYEGQGSJHTVVY4WAFAQSFL","short_pith_number":"pith:LO5BPYEG","schema_version":"1.0","canonical_sha256":"5bba17e08681a493ceb5c72c0282122ae8dc119b204e9c2dc78ae592a76f2ad5","source":{"kind":"arxiv","id":"2605.16745","version":1},"attestation_state":"computed","paper":{"title":"EVA01: Unified Native 3D Understanding and Generation via Mixture-of-Transformers","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"EVA01 integrates 3D meshes as a native modality inside multimodal language models using a mixture-of-transformers split.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Baolin Liu, Bocheng Li, Chenzhuo Fan, Mingjing Yi, Shimu Wang, Wanli Ma, Yingde Song, Yongping Xiong, Yuke Lou, Zhengdong Guo, Zongyuan Yang","submitted_at":"2026-05-16T01:55:03Z","abstract_excerpt":"This paper addresses the challenge of integrating 3D meshes as a native modality within Multimodal Large Language Models (MLLMs). Diffusion-based large reconstruction models decouple semantic understanding from geometric reasoning, operating as stateless reconstructors conditioned on dense 2D pixel priors. Recent MLLM-based methods treat the 3D modality as an external output rather than a native component of the multimodal sequence, making incremental adaptations without a systematic analysis of how geometric manifolds align with MLLM feature spaces. We introduce EVA01, a unified framework tha"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.16745","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-16T01:55:03Z","cross_cats_sorted":[],"title_canon_sha256":"e81542ff15bc1f6f5509b3fe4842b5b48c686b3d1b03fed83f8524096c3ede09","abstract_canon_sha256":"499a72a158fbe86ee84d8621bbbb168bf7da06e44a56bf595a23c08d3c99c033"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:39.528330Z","signature_b64":"lgVeEEHzondM3P4YiBLqFSo2YoQKxv4iYGWsVYrIu7RsJuF3lohZ2XP1uKEbNMGTRM/E1fnoGM1agObwgwEdCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5bba17e08681a493ceb5c72c0282122ae8dc119b204e9c2dc78ae592a76f2ad5","last_reissued_at":"2026-05-20T00:02:39.527485Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:39.527485Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"EVA01: Unified Native 3D Understanding and Generation via Mixture-of-Transformers","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"EVA01 integrates 3D meshes as a native modality inside multimodal language models using a mixture-of-transformers split.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Baolin Liu, Bocheng Li, Chenzhuo Fan, Mingjing Yi, Shimu Wang, Wanli Ma, Yingde Song, Yongping Xiong, Yuke Lou, Zhengdong Guo, Zongyuan Yang","submitted_at":"2026-05-16T01:55:03Z","abstract_excerpt":"This paper addresses the challenge of integrating 3D meshes as a native modality within Multimodal Large Language Models (MLLMs). Diffusion-based large reconstruction models decouple semantic understanding from geometric reasoning, operating as stateless reconstructors conditioned on dense 2D pixel priors. Recent MLLM-based methods treat the 3D modality as an external output rather than a native component of the multimodal sequence, making incremental adaptations without a systematic analysis of how geometric manifolds align with MLLM feature spaces. We introduce EVA01, a unified framework tha"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"EVA01 achieves state-of-the-art native text-to-3D generation fidelity and unlocks robust long-context multi-turn geometric editing with identity preservation, a capability fundamentally inaccessible to stateless reconstruction pipelines.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumption that decoupling into a pre-trained Understanding Expert and a structurally mirrored Generation Expert, coupled through shared global self-attention with hard modality routing, will align the semantic latent space of the MLLM backbone with the geometric manifold without performance loss or the need for intermediate 2D representations.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"EVA01 introduces a Mixture-of-Transformers model that natively adds 3D mesh understanding, generation, and multi-turn editing to MLLMs by decoupling understanding and generation experts with shared global self-attention.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"EVA01 integrates 3D meshes as a native modality inside multimodal language models using a mixture-of-transformers split.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ac27ce6371b0a894f587382b295671f811bff0dd8bcdfe642790e121eae87902"},"source":{"id":"2605.16745","kind":"arxiv","version":1},"verdict":{"id":"eed022b1-7a92-4f7e-8f74-0fc019656e51","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T21:20:55.777240Z","strongest_claim":"EVA01 achieves state-of-the-art native text-to-3D generation fidelity and unlocks robust long-context multi-turn geometric editing with identity preservation, a capability fundamentally inaccessible to stateless reconstruction pipelines.","one_line_summary":"EVA01 introduces a Mixture-of-Transformers model that natively adds 3D mesh understanding, generation, and multi-turn editing to MLLMs by decoupling understanding and generation experts with shared global self-attention.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumption that decoupling into a pre-trained Understanding Expert and a structurally mirrored Generation Expert, coupled through shared global self-attention with hard modality routing, will align the semantic latent space of the MLLM backbone with the geometric manifold without performance loss or the need for intermediate 2D representations.","pith_extraction_headline":"EVA01 integrates 3D meshes as a native modality inside multimodal language models using a mixture-of-transformers split."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.16745/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T21:31:19.388825Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T21:31:15.256218Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T19:01:56.331028Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T18:33:26.461039Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"5096c211e3a6740ba9a11abcfee541ee07dc279d59b470125d8341bd3b1d120f"},"references":{"count":77,"sample":[{"doi":"","year":2025,"title":"Qwen3-VL Technical Report","work_id":"1fe243aa-e3c0-4da6-b391-4cbcfc88d5c0","ref_index":1,"cited_arxiv_id":"2511.21631","is_internal_anchor":true},{"doi":"","year":2005,"title":"METEOR : An automatic metric for MT evaluation with improved correlation with human judgments","work_id":"8ef377b6-8e72-424b-9cca-582931404df4","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Instant3DiT : Multiview inpainting for fast editing of 3D objects","work_id":"7870fecc-1a60-4a03-82c4-67dde070210a","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2026,"title":"TIPSv2: Advancing Vision-Language Pretraining with Enhanced Patch-Text Alignment","work_id":"5e9b3ddf-5b6b-4042-b640-d0d8757bfe14","ref_index":4,"cited_arxiv_id":"2604.12012","is_internal_anchor":true},{"doi":"","year":1918,"title":"ShapeNet: An Information-Rich 3D Model Repository","work_id":"b2ac5b60-daa9-435b-9369-12271e126edd","ref_index":5,"cited_arxiv_id":"1512.03012","is_internal_anchor":true}],"resolved_work":77,"snapshot_sha256":"b15689ee741e1a9089a564fef7de8ea19ac3b7209e0f83b88389668c9f9100c3","internal_anchors":13},"formal_canon":{"evidence_count":2,"snapshot_sha256":"f73224ace45fc37e41361f5897d8e99e8833124e1119b6acea0027bcc3348e21"},"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":"2605.16745","created_at":"2026-05-20T00:02:39.527619+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.16745v1","created_at":"2026-05-20T00:02:39.527619+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.16745","created_at":"2026-05-20T00:02:39.527619+00:00"},{"alias_kind":"pith_short_12","alias_value":"LO5BPYEGQGSJ","created_at":"2026-05-20T00:02:39.527619+00:00"},{"alias_kind":"pith_short_16","alias_value":"LO5BPYEGQGSJHTVV","created_at":"2026-05-20T00:02:39.527619+00:00"},{"alias_kind":"pith_short_8","alias_value":"LO5BPYEG","created_at":"2026-05-20T00:02:39.527619+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL","json":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL.json","graph_json":"https://pith.science/api/pith-number/LO5BPYEGQGSJHTVVY4WAFAQSFL/graph.json","events_json":"https://pith.science/api/pith-number/LO5BPYEGQGSJHTVVY4WAFAQSFL/events.json","paper":"https://pith.science/paper/LO5BPYEG"},"agent_actions":{"view_html":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL","download_json":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL.json","view_paper":"https://pith.science/paper/LO5BPYEG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.16745&json=true","fetch_graph":"https://pith.science/api/pith-number/LO5BPYEGQGSJHTVVY4WAFAQSFL/graph.json","fetch_events":"https://pith.science/api/pith-number/LO5BPYEGQGSJHTVVY4WAFAQSFL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL/action/storage_attestation","attest_author":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL/action/author_attestation","sign_citation":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL/action/citation_signature","submit_replication":"https://pith.science/pith/LO5BPYEGQGSJHTVVY4WAFAQSFL/action/replication_record"}},"created_at":"2026-05-20T00:02:39.527619+00:00","updated_at":"2026-05-20T00:02:39.527619+00:00"}