[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84469-en":3,"doc-seo-84469-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},84469,687197100911,"Himbo","https://ap-avatar.wpscdn.com/avatar/a000239b6f1da00475?x-image-process=image/resize,m_fixed,w_180,h_180&k=1782698725881665579",8,"Research & Report","Local Message-Passing for Discrete Graph Generation","Discrete graph generation models graph-structured data, but leading approaches commonly assume Graph Transformers or higher-order architectures. GenGNN is proposed as a modular local message-passing backbone for discrete graph generation, persisting edge fields via latent refinement of coupled node–edge–graph states without global attention. Diffusion models integrating GenGNN reach over 90% validity on standard benchmarks, matching Graph Transformer backbones and running 2–5× faster inference. Ablations attribute robustness to components that mitigate oversmoothing, and representation analysis shows functionally similar learned representations even at deeper layers.","Local Message-Passing for Discrete Graph Generation  \nJay Revolinsky  \n[revolins@msu.edu](revolins@msu.edu)[ ](revolins@msu.edu)Michigan State University East Lansing, Michigan, USA  \nHarry Shomer  \nUniversity of Texas-Arlington Arlington, Texas, USA  \nJiliang Tang  \nMichigan State University East Lansing, Michigan, USA  \narXiv :2603 .08825v2 [ cs .LG] 12 Jul 2026  \nAbstract  \nDiscrete graph generation has emerged as a powerful paradigm for modeling graph-structured data, yet state-of-the-art models often rely on Graph Transformers or higher-order architectures. We revisit this design assumption by introducing GenGNN, a modular message-passing backbone for graph generation. GenGNN enables powerful generation by persisting edge fields through latent refinement of coupled node–edge–graph states, all without requiring global attention. Diffusion models integrating GenGNN achieve over 90% validity on standard benchmark datasets, performing within margins of Graph Transformer backbones and achieving 2–5x faster inference. Systematic ablations isolate how GenGNN is resilient to oversmoothing during generative (de)noising, indicating each GenGNN component is necessary for downstream generation quality. Finally, representation-space analysis suggests GenGNN learns functionally-similar representations to more theoreticallyexpressive architectures; even at deeper layers. As such, GenGNN uplifts local message-passing to challenge prevailing assumptions that performant discrete graph generation requires global attention or higher-order representations. Source Code: Available Here  \nCCS Concepts  \n• Computing methodologies → Neural networks; • Theory of computation → Graph algorithms analysis.  \nKeywords  \nGraph Generation, Message-Passing Neural Networks, Graph Neural Networks, Discrete Flow-Matching, Discrete Diffusion  \nACM Reference Format:  \nJay Revolinsky, Harry Shomer, and Jiliang Tang. 2026. Local MessagePassing for Discrete Graph Generation. In Proceedings of Conference on Knowledge and Information Management (CIKM’26). ACM, New York, NY, USA, 11 pages. [https://doi.org/XXXXXXX.XXXXXXX](https://doi.org/XXXXXXX.XXXXXXX)  \n1 Introduction  \nGraph generation is a rapidly growing subfield of graph representation learning, with applications to drug discovery [53] and code modeling [5] . Recent graph generative models (GGMs) based on discrete diffusion achieve strong downstream performance by iteratively denoising discrete node and edge states [14, 48] . More  \nPermission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission [and/or a fee. Request permissions from permissions@acm.org](and/or a fee. Request permissions from permissions@acm.org).  \nCIKM’26, Rome, Italy  \n© 2026 Copyright held by the owner/author(s) . Publication rights licensed to ACM. ACM ISBN 978-1-4503-XXXX-X/2018/06  \n[https://doi.org/XXXXXXX.XXXXXXX](https://doi.org/XXXXXXX.XXXXXXX)  \n(a) Simple GNN (b) Graph Transformer  \nFigure 1: Planar graphs generated via DeFoG with either a simple GNN or Graph Transformer backbone. The simple GNN fails to reproduce finer substructures in its output.  \nrecent work further integrates continuous-time flow modeling into discrete graph generation [42, 54], resulting in strong performance on standard benchmark datasets [36] . However, this performance comes with a substantial inference cost, as discrete graph generation typically requires repeated denoising steps over dense node–edge representations which are quadratic in complexity [36, 48] .  \nExisting work has primarily addressed","cbCaijN6MSEMXIVX","https://ap.wps.com/l/cbCaijN6MSEMXIVX","pdf",946032,1,11,"English","en",105,"# Abstract\n# Introduction\n## Motivation and Background\n## Computational Bottlenecks in Discrete Graph Generation\n## Limits of Local vs Non-Local Backbones","[{\"question\":\"What is GenGNN and what problem does it address in discrete graph generation?\",\"answer\":\"GenGNN is a modular message-passing backbone that generates discrete graphs by refining coupled node–edge–graph states. It targets the cost and complexity of expensive denoising backbones often used in state-of-the-art discrete diffusion graph models.\"},{\"question\":\"How does GenGNN avoid relying on global attention or higher-order architectures?\",\"answer\":\"GenGNN persists edge fields through latent refinement of coupled node–edge–graph states while using local message passing rather than global attention mechanisms. This enables diffusion-based generation without transformer-style non-local processing.\"},{\"question\":\"What performance and efficiency gains are reported for diffusion models that integrate GenGNN?\",\"answer\":\"Integrating GenGNN into diffusion models achieves over 90% validity on standard benchmark datasets. Reported inference is within the margin of Graph Transformer backbones while being 2–5× faster during inference.\"}]",1784195841,28,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"local-message-passing-for-discrete-graph-generation","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/local-message-passing-for-discrete-graph-generation/84469/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is GenGNN and what problem does it address in discrete graph generation?","Question",{"text":75,"@type":76},"GenGNN is a modular message-passing backbone that generates discrete graphs by refining coupled node–edge–graph states. It targets the cost and complexity of expensive denoising backbones often used in state-of-the-art discrete diffusion graph models.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does GenGNN avoid relying on global attention or higher-order architectures?",{"text":80,"@type":76},"GenGNN persists edge fields through latent refinement of coupled node–edge–graph states while using local message passing rather than global attention mechanisms. This enables diffusion-based generation without transformer-style non-local processing.",{"name":82,"@type":73,"acceptedAnswer":83},"What performance and efficiency gains are reported for diffusion models that integrate GenGNN?",{"text":84,"@type":76},"Integrating GenGNN into diffusion models achieves over 90% validity on standard benchmark datasets. Reported inference is within the margin of Graph Transformer backbones while being 2–5× faster during inference.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]