{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6GWP6UXDC5DV4LQY24DP3QRZAX","short_pith_number":"pith:6GWP6UXD","schema_version":"1.0","canonical_sha256":"f1acff52e317475e2e18d706fdc23905e62400e16a1bb1a3a5dd9cf81ae890a7","source":{"kind":"arxiv","id":"2412.03069","version":2},"attestation_state":"computed","paper":{"title":"TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Daniel K. Du, Huichao Zhang, Hu Ye, Liao Qu, Xinglong Wu, Xu Wang, Yiheng Liu, Yi Jiang, Yiming Gao, Zehuan Yuan","submitted_at":"2024-12-04T06:46:55Z","abstract_excerpt":"We present TokenFlow, a novel unified image tokenizer that bridges the long-standing gap between multimodal understanding and generation. Prior research attempt to employ a single reconstruction-targeted Vector Quantization (VQ) encoder for unifying these two tasks. We observe that understanding and generation require fundamentally different granularities of visual information. This leads to a critical trade-off, particularly compromising performance in multimodal understanding tasks. TokenFlow addresses this challenge through an innovative dual-codebook architecture that decouples semantic an"},"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":"2412.03069","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-12-04T06:46:55Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"546d6de98e7823d0cf1656ba78ba9826a8d190c02780b9a1acddbe882068c839","abstract_canon_sha256":"fbdc3d3030e2b1daa756daeef0817e7af17c1fd6456a7dd5a4094fe6b33cff29"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:49:48.461690Z","signature_b64":"b5GydUTelCCQOq/P3XYsFW9Wp6W8cMaZ9gMKHhx1vYn26Pb+scYfHsSKhTRTWFXTm2sFeGGApBMZITA//CTPCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f1acff52e317475e2e18d706fdc23905e62400e16a1bb1a3a5dd9cf81ae890a7","last_reissued_at":"2026-07-05T11:49:48.461219Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:49:48.461219Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Daniel K. Du, Huichao Zhang, Hu Ye, Liao Qu, Xinglong Wu, Xu Wang, Yiheng Liu, Yi Jiang, Yiming Gao, Zehuan Yuan","submitted_at":"2024-12-04T06:46:55Z","abstract_excerpt":"We present TokenFlow, a novel unified image tokenizer that bridges the long-standing gap between multimodal understanding and generation. Prior research attempt to employ a single reconstruction-targeted Vector Quantization (VQ) encoder for unifying these two tasks. We observe that understanding and generation require fundamentally different granularities of visual information. This leads to a critical trade-off, particularly compromising performance in multimodal understanding tasks. TokenFlow addresses this challenge through an innovative dual-codebook architecture that decouples semantic an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.03069","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/2412.03069/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":"2412.03069","created_at":"2026-07-05T11:49:48.461274+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.03069v2","created_at":"2026-07-05T11:49:48.461274+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.03069","created_at":"2026-07-05T11:49:48.461274+00:00"},{"alias_kind":"pith_short_12","alias_value":"6GWP6UXDC5DV","created_at":"2026-07-05T11:49:48.461274+00:00"},{"alias_kind":"pith_short_16","alias_value":"6GWP6UXDC5DV4LQY","created_at":"2026-07-05T11:49:48.461274+00:00"},{"alias_kind":"pith_short_8","alias_value":"6GWP6UXD","created_at":"2026-07-05T11:49:48.461274+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":15,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.31683","citing_title":"Histogram-constrained Image Generation","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2503.14324","citing_title":"DualToken: Towards Unifying Visual Understanding and Generation with Dual Visual Vocabularies","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18390","citing_title":"Vision Foundation Models as Generalist Tokenizers for Image Generation","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2505.05472","citing_title":"Mogao: An Omni Foundation Model for Interleaved Multi-Modal Generation","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2601.15507","citing_title":"A Unified and Controllable Framework for Layered Image Generation with Visual Effects","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03893","citing_title":"FeynmanBench: Benchmarking Multimodal LLMs on Diagrammatic Physics Reasoning","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2506.15564","citing_title":"Show-o2: Improved Native Unified Multimodal Models","ref_index":89,"is_internal_anchor":false},{"citing_arxiv_id":"2604.28185","citing_title":"Visual Generation in the New Era: An Evolution from Atomic Mapping to Agentic World Modeling","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2505.09568","citing_title":"BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24625","citing_title":"Meta-CoT: Enhancing Granularity and Generalization in Image Editing","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2501.17811","citing_title":"Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07915","citing_title":"What Matters for Diffusion-Friendly Latent Manifold? Prior-Aligned Autoencoders for Latent Diffusion","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04746","citing_title":"Think in Strokes, Not Pixels: Process-Driven Image Generation via Interleaved Reasoning","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2505.14683","citing_title":"Emerging Properties in Unified Multimodal Pretraining","ref_index":59,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13540","citing_title":"Free Lunch for Unified Multimodal Models: Enhancing Generation via Reflective Rectification with Inherent Understanding","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX","json":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX.json","graph_json":"https://pith.science/api/pith-number/6GWP6UXDC5DV4LQY24DP3QRZAX/graph.json","events_json":"https://pith.science/api/pith-number/6GWP6UXDC5DV4LQY24DP3QRZAX/events.json","paper":"https://pith.science/paper/6GWP6UXD"},"agent_actions":{"view_html":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX","download_json":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX.json","view_paper":"https://pith.science/paper/6GWP6UXD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.03069&json=true","fetch_graph":"https://pith.science/api/pith-number/6GWP6UXDC5DV4LQY24DP3QRZAX/graph.json","fetch_events":"https://pith.science/api/pith-number/6GWP6UXDC5DV4LQY24DP3QRZAX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX/action/storage_attestation","attest_author":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX/action/author_attestation","sign_citation":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX/action/citation_signature","submit_replication":"https://pith.science/pith/6GWP6UXDC5DV4LQY24DP3QRZAX/action/replication_record"}},"created_at":"2026-07-05T11:49:48.461274+00:00","updated_at":"2026-07-05T11:49:48.461274+00:00"}