{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VBTBECI43ZLD4XJICIQQU46PZX","short_pith_number":"pith:VBTBECI4","schema_version":"1.0","canonical_sha256":"a86612091cde563e5d2812210a73cfcdf08fd236a37537eb4b2877df9e556ca5","source":{"kind":"arxiv","id":"2604.13030","version":2},"attestation_state":"computed","paper":{"title":"Generative Refinement Networks for Visual Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Generative Refinement Networks combine near-lossless quantization with global refinement to surpass diffusion and autoregressive models in visual synthesis.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bingyue Peng, Jiahuan Wang, Jian Han, Jinlai Liu, Zehuan Yuan","submitted_at":"2026-04-14T17:59:03Z","abstract_excerpt":"While diffusion models dominate the field of visual generation, they are computationally inefficient, applying a uniform computational effort regardless of different complexity. In contrast, autoregressive (AR) models are inherently complexity-aware, as evidenced by their variable likelihoods, but are often hindered by lossy discrete tokenization and error accumulation. In this work, we introduce Generative Refinement Networks (GRN), a next-generation visual synthesis paradigm that addresses these issues. At its core, GRN addresses the discrete tokenization bottleneck through a theoretically n"},"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":"2604.13030","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-04-14T17:59:03Z","cross_cats_sorted":[],"title_canon_sha256":"882a87b9ada1a0488dcca7fa5e5ec821c4a407cfbb87e48990287d81c6e2d0f4","abstract_canon_sha256":"977f9bb339fa6c6087184224f4f8e59ed2abb726d43fe819f15aecea94e8ea88"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-08T01:19:12.859752Z","signature_b64":"x8tuQMmsvPvaBqG8zO5LkCIhyTr4S/NvnaQpLs4e9MqJ7obGPq+rHPtm04p5JS5q25kE/flAHw1nGGtX8wUGAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a86612091cde563e5d2812210a73cfcdf08fd236a37537eb4b2877df9e556ca5","last_reissued_at":"2026-07-08T01:19:12.859218Z","signature_status":"signed_v1","first_computed_at":"2026-07-08T01:19:12.859218Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generative Refinement Networks for Visual Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Generative Refinement Networks combine near-lossless quantization with global refinement to surpass diffusion and autoregressive models in visual synthesis.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bingyue Peng, Jiahuan Wang, Jian Han, Jinlai Liu, Zehuan Yuan","submitted_at":"2026-04-14T17:59:03Z","abstract_excerpt":"While diffusion models dominate the field of visual generation, they are computationally inefficient, applying a uniform computational effort regardless of different complexity. In contrast, autoregressive (AR) models are inherently complexity-aware, as evidenced by their variable likelihoods, but are often hindered by lossy discrete tokenization and error accumulation. In this work, we introduce Generative Refinement Networks (GRN), a next-generation visual synthesis paradigm that addresses these issues. At its core, GRN addresses the discrete tokenization bottleneck through a theoretically n"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"On the ImageNet benchmark, GRN establishes new records in image reconstruction (0.56 rFID) and class-conditional image generation (1.81 gFID). We also scale GRN to more challenging text-to-image and text-to-video generation, delivering superior performance on an equivalent scale.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the Hierarchical Binary Quantization is theoretically near-lossless and that the global refinement mechanism corrects errors without introducing new accumulation problems or requiring post-hoc tuning that affects the reported metrics.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"GRN uses hierarchical binary quantization and entropy-guided refinement to set new ImageNet records of 0.56 rFID for reconstruction and 1.81 gFID for class-conditional generation while releasing code and models.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Generative Refinement Networks combine near-lossless quantization with global refinement to surpass diffusion and autoregressive models in visual synthesis.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7237182ebe3a0838a0eb5990bbc9ea321f13487929f1ffe114e96713404a2a84"},"source":{"id":"2604.13030","kind":"arxiv","version":2},"verdict":{"id":"b00fa9a5-57b3-449e-a0aa-8aa439dde081","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T15:37:19.049077Z","strongest_claim":"On the ImageNet benchmark, GRN establishes new records in image reconstruction (0.56 rFID) and class-conditional image generation (1.81 gFID). We also scale GRN to more challenging text-to-image and text-to-video generation, delivering superior performance on an equivalent scale.","one_line_summary":"GRN uses hierarchical binary quantization and entropy-guided refinement to set new ImageNet records of 0.56 rFID for reconstruction and 1.81 gFID for class-conditional generation while releasing code and models.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the Hierarchical Binary Quantization is theoretically near-lossless and that the global refinement mechanism corrects errors without introducing new accumulation problems or requiring post-hoc tuning that affects the reported metrics.","pith_extraction_headline":"Generative Refinement Networks combine near-lossless quantization with global refinement to surpass diffusion and autoregressive models in visual synthesis."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.13030/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":"2604.13030","created_at":"2026-07-08T01:19:12.859276+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.13030v2","created_at":"2026-07-08T01:19:12.859276+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.13030","created_at":"2026-07-08T01:19:12.859276+00:00"},{"alias_kind":"pith_short_12","alias_value":"VBTBECI43ZLD","created_at":"2026-07-08T01:19:12.859276+00:00"},{"alias_kind":"pith_short_16","alias_value":"VBTBECI43ZLD4XJI","created_at":"2026-07-08T01:19:12.859276+00:00"},{"alias_kind":"pith_short_8","alias_value":"VBTBECI4","created_at":"2026-07-08T01:19:12.859276+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.00562","citing_title":"Beyond Token-Level Cross-Entropy: Fr\\'echet Distributional Post-Training for Autoregressive Image Generation","ref_index":47,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX","json":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX.json","graph_json":"https://pith.science/api/pith-number/VBTBECI43ZLD4XJICIQQU46PZX/graph.json","events_json":"https://pith.science/api/pith-number/VBTBECI43ZLD4XJICIQQU46PZX/events.json","paper":"https://pith.science/paper/VBTBECI4"},"agent_actions":{"view_html":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX","download_json":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX.json","view_paper":"https://pith.science/paper/VBTBECI4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.13030&json=true","fetch_graph":"https://pith.science/api/pith-number/VBTBECI43ZLD4XJICIQQU46PZX/graph.json","fetch_events":"https://pith.science/api/pith-number/VBTBECI43ZLD4XJICIQQU46PZX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX/action/storage_attestation","attest_author":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX/action/author_attestation","sign_citation":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX/action/citation_signature","submit_replication":"https://pith.science/pith/VBTBECI43ZLD4XJICIQQU46PZX/action/replication_record"}},"created_at":"2026-07-08T01:19:12.859276+00:00","updated_at":"2026-07-08T01:19:12.859276+00:00"}