{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:LO5BPYEGQGSJHTVVY4WAFAQSFL","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"499a72a158fbe86ee84d8621bbbb168bf7da06e44a56bf595a23c08d3c99c033","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-16T01:55:03Z","title_canon_sha256":"e81542ff15bc1f6f5509b3fe4842b5b48c686b3d1b03fed83f8524096c3ede09"},"schema_version":"1.0","source":{"id":"2605.16745","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.16745","created_at":"2026-05-20T00:02:39Z"},{"alias_kind":"arxiv_version","alias_value":"2605.16745v1","created_at":"2026-05-20T00:02:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.16745","created_at":"2026-05-20T00:02:39Z"},{"alias_kind":"pith_short_12","alias_value":"LO5BPYEGQGSJ","created_at":"2026-05-20T00:02:39Z"},{"alias_kind":"pith_short_16","alias_value":"LO5BPYEGQGSJHTVV","created_at":"2026-05-20T00:02:39Z"},{"alias_kind":"pith_short_8","alias_value":"LO5BPYEG","created_at":"2026-05-20T00:02:39Z"}],"graph_snapshots":[{"event_id":"sha256:16fe8b5cf65748a66526a374b2d0b6a9a34bda57ebfe077956e4da62a0bcd344","target":"graph","created_at":"2026-05-20T00:02:39Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","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."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","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."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","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."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"EVA01 integrates 3D meshes as a native modality inside multimodal language models using a mixture-of-transformers split."}],"snapshot_sha256":"ac27ce6371b0a894f587382b295671f811bff0dd8bcdfe642790e121eae87902"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"f73224ace45fc37e41361f5897d8e99e8833124e1119b6acea0027bcc3348e21"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"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"}],"endpoint":"/pith/2605.16745/integrity.json","findings":[],"snapshot_sha256":"5096c211e3a6740ba9a11abcfee541ee07dc279d59b470125d8341bd3b1d120f","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"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","authors_text":"Baolin Liu, Bocheng Li, Chenzhuo Fan, Mingjing Yi, Shimu Wang, Wanli Ma, Yingde Song, Yongping Xiong, Yuke Lou, Zhengdong Guo, Zongyuan Yang","cross_cats":[],"headline":"EVA01 integrates 3D meshes as a native modality inside multimodal language models using a mixture-of-transformers split.","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-16T01:55:03Z","title":"EVA01: Unified Native 3D Understanding and Generation via Mixture-of-Transformers"},"references":{"count":77,"internal_anchors":13,"resolved_work":77,"sample":[{"cited_arxiv_id":"2511.21631","doi":"","is_internal_anchor":true,"ref_index":1,"title":"Qwen3-VL Technical Report","work_id":"1fe243aa-e3c0-4da6-b391-4cbcfc88d5c0","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"METEOR : An automatic metric for MT evaluation with improved correlation with human judgments","work_id":"8ef377b6-8e72-424b-9cca-582931404df4","year":2005},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Instant3DiT : Multiview inpainting for fast editing of 3D objects","work_id":"7870fecc-1a60-4a03-82c4-67dde070210a","year":2025},{"cited_arxiv_id":"2604.12012","doi":"","is_internal_anchor":true,"ref_index":4,"title":"TIPSv2: Advancing Vision-Language Pretraining with Enhanced Patch-Text Alignment","work_id":"5e9b3ddf-5b6b-4042-b640-d0d8757bfe14","year":2026},{"cited_arxiv_id":"1512.03012","doi":"","is_internal_anchor":true,"ref_index":5,"title":"ShapeNet: An Information-Rich 3D Model Repository","work_id":"b2ac5b60-daa9-435b-9369-12271e126edd","year":1918}],"snapshot_sha256":"b15689ee741e1a9089a564fef7de8ea19ac3b7209e0f83b88389668c9f9100c3"},"source":{"id":"2605.16745","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-19T21:20:55.777240Z","id":"eed022b1-7a92-4f7e-8f74-0fc019656e51","model_set":{"reader":"grok-4.3"},"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","pith_extraction_headline":"EVA01 integrates 3D meshes as a native modality inside multimodal language models using a mixture-of-transformers split.","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.","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."}},"verdict_id":"eed022b1-7a92-4f7e-8f74-0fc019656e51"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:cc10912e7efc6314580f32a73cb1cdd7f1fcf3f1e8583d52d28346834142a37a","target":"record","created_at":"2026-05-20T00:02:39Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"499a72a158fbe86ee84d8621bbbb168bf7da06e44a56bf595a23c08d3c99c033","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-16T01:55:03Z","title_canon_sha256":"e81542ff15bc1f6f5509b3fe4842b5b48c686b3d1b03fed83f8524096c3ede09"},"schema_version":"1.0","source":{"id":"2605.16745","kind":"arxiv","version":1}},"canonical_sha256":"5bba17e08681a493ceb5c72c0282122ae8dc119b204e9c2dc78ae592a76f2ad5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5bba17e08681a493ceb5c72c0282122ae8dc119b204e9c2dc78ae592a76f2ad5","first_computed_at":"2026-05-20T00:02:39.527485Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:02:39.527485Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lgVeEEHzondM3P4YiBLqFSo2YoQKxv4iYGWsVYrIu7RsJuF3lohZ2XP1uKEbNMGTRM/E1fnoGM1agObwgwEdCA==","signature_status":"signed_v1","signed_at":"2026-05-20T00:02:39.528330Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.16745","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:cc10912e7efc6314580f32a73cb1cdd7f1fcf3f1e8583d52d28346834142a37a","sha256:16fe8b5cf65748a66526a374b2d0b6a9a34bda57ebfe077956e4da62a0bcd344"],"state_sha256":"2c3fbf30f58018d01f67eb71174c3ee57b49cb37efaa0cdb573c648875ea6d8d"}