{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:6E2K2UBHSY36A36SUXX4CLCRCK","short_pith_number":"pith:6E2K2UBH","schema_version":"1.0","canonical_sha256":"f134ad50279637e06fd2a5efc12c511287ad8990f8f4c0a30e11a04962608ac7","source":{"kind":"arxiv","id":"2504.21356","version":3},"attestation_state":"computed","paper":{"title":"Nexus-Gen: Unified Image Understanding, Generation, and Editing via Prefilled Autoregression in Shared Embedding Space","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Hong Zhang, Weiyi Lu, Xingjun Wang, Yingda Chen, Yixuan Xu, Yuze Zhao, Yu Zhang, Zhipeng Di, Zhongjie Duan","submitted_at":"2025-04-30T06:30:48Z","abstract_excerpt":"Unified multimodal generative models aim to integrate image understanding and generation abilities, offering significant advantages in harnessing multimodal corpora, particularly interleaved text-image data. However, existing unified models exhibit limitations in image synthesis quality, autoregressive error accumulation, and image editing capability. In this work, we propose Nexus-Gen, a novel architecture that unifies image understanding, generation, and editing tasks in a shared image embedding space. This shared space serves as a bridge for the autoregressive and diffusion models, which se"},"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":"2504.21356","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-04-30T06:30:48Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"d11eb176fb4b6067dac279a5ca6c609466328e0665f1d048f5b4405735286443","abstract_canon_sha256":"7bd22b36cdc9f359b3607c84f3822dcda1ede31b7346f27f1eed28ed7457c723"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:19.058546Z","signature_b64":"m68vfZDzsVc2WvlvBMlgMwBR93l153xmQI8tncC4j0tH0fcRmbBpPdRFuX2Rt2hLSoGTfpw9rZWmF/yRVBnECg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f134ad50279637e06fd2a5efc12c511287ad8990f8f4c0a30e11a04962608ac7","last_reissued_at":"2026-07-05T11:37:19.058053Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:19.058053Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Nexus-Gen: Unified Image Understanding, Generation, and Editing via Prefilled Autoregression in Shared Embedding Space","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Hong Zhang, Weiyi Lu, Xingjun Wang, Yingda Chen, Yixuan Xu, Yuze Zhao, Yu Zhang, Zhipeng Di, Zhongjie Duan","submitted_at":"2025-04-30T06:30:48Z","abstract_excerpt":"Unified multimodal generative models aim to integrate image understanding and generation abilities, offering significant advantages in harnessing multimodal corpora, particularly interleaved text-image data. However, existing unified models exhibit limitations in image synthesis quality, autoregressive error accumulation, and image editing capability. In this work, we propose Nexus-Gen, a novel architecture that unifies image understanding, generation, and editing tasks in a shared image embedding space. This shared space serves as a bridge for the autoregressive and diffusion models, which se"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.21356","kind":"arxiv","version":3},"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/2504.21356/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":"2504.21356","created_at":"2026-07-05T11:37:19.058110+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.21356v3","created_at":"2026-07-05T11:37:19.058110+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.21356","created_at":"2026-07-05T11:37:19.058110+00:00"},{"alias_kind":"pith_short_12","alias_value":"6E2K2UBHSY36","created_at":"2026-07-05T11:37:19.058110+00:00"},{"alias_kind":"pith_short_16","alias_value":"6E2K2UBHSY36A36S","created_at":"2026-07-05T11:37:19.058110+00:00"},{"alias_kind":"pith_short_8","alias_value":"6E2K2UBH","created_at":"2026-07-05T11:37:19.058110+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23835","citing_title":"ABACUS: Adapting Unified Foundation Model for Bridging Image Count Understanding and Generation","ref_index":90,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13289","citing_title":"HYDRA-X: Native Unified Multimodal Models with Holistic Visual Tokenizers","ref_index":107,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05071","citing_title":"InstantRetouch: Efficient and High-Fidelity Instruction-Guided Image Retouching with Bilateral Space","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2506.15564","citing_title":"Show-o2: Improved Native Unified Multimodal Models","ref_index":141,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06870","citing_title":"RefineAnything: Multimodal Region-Specific Refinement for Perfect Local Details","ref_index":52,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK","json":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK.json","graph_json":"https://pith.science/api/pith-number/6E2K2UBHSY36A36SUXX4CLCRCK/graph.json","events_json":"https://pith.science/api/pith-number/6E2K2UBHSY36A36SUXX4CLCRCK/events.json","paper":"https://pith.science/paper/6E2K2UBH"},"agent_actions":{"view_html":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK","download_json":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK.json","view_paper":"https://pith.science/paper/6E2K2UBH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.21356&json=true","fetch_graph":"https://pith.science/api/pith-number/6E2K2UBHSY36A36SUXX4CLCRCK/graph.json","fetch_events":"https://pith.science/api/pith-number/6E2K2UBHSY36A36SUXX4CLCRCK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK/action/storage_attestation","attest_author":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK/action/author_attestation","sign_citation":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK/action/citation_signature","submit_replication":"https://pith.science/pith/6E2K2UBHSY36A36SUXX4CLCRCK/action/replication_record"}},"created_at":"2026-07-05T11:37:19.058110+00:00","updated_at":"2026-07-05T11:37:19.058110+00:00"}