[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81933-en":3,"doc-seo-81933-105":29,"detail-sidebar-cat-0-en-105":90},{"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":4,"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},81933,8796095461564,"Liam","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Parameter-Free Encoders Remain Viable for RDB Foundation Models","Relational database (RDB) foundation models aim to predict missing or future values in a target column across heterogeneous enterprise tables without retraining a new model for every task. Frozen RDB-specific encoders are a practical option, yet encoder design is debated: parameter-free subgraph encoders with single-table foundation models can approach near SOTA, while parameterized encoders pre-trained with observable labels claim task-specific representational gains. This work analyzes label-driven encoder roles and proves limitations for trainable encoder parameters, supported by empirical results showing simpler parameter-free encoders remain effective across benchmarks.","Parameter-Free Encoders Remain Viable for RDB Foundation Models  \nLinjie Xu 1 David Wipf 1  \narXiv :2607 .05476v 1 [ cs .LG] 6 Jul 2026  \nAbstract  \nGiven a relational database (RDB) storing heterogeneous tabular information, how can we predict missing (or future) values in some target column of interest? As the space of potential targets is vast across enterprise settings, it is preferable to avoid learning a new model from scratch each time there is a new prediction task. Frozen foundation models based on RDB-specific encoders provide a viable solution, but ideal design remains an open question. On the one hand, it has recently been argued that certain parameter-free subgraph encoders combined with single-table foundation models can achieve near SOTA performance, with no RDB-specific pre-training required. Meanwhile, other contemporary studies advocate for parameterized encoders pre-trained to exploit observable labels for learning task-specific representations. To address this ambiguity, we analyze RDB encoder properties specifically when labels are present as inputs, proving limitations on the potential efficacy of trainable encoder parameters. As empirical validation, we demonstrate that considerably simpler parameter-free encoders are still capable of strong performance across many relevant benchmarking tasks.  \n1. Introduction  \nRelational databases (RDBs), housing collections of interrelated tables, are indispensable across wide-ranging enterprise applications and e-commerce platforms (GarciaMolina et al., 2009) . From a machine learning standpoint, the information stored within just a single RDB often contains countless possibilities for predictive modeling, covering targets such as customer retention, user churn, or clickthrough rates (Dave et al., 2014 ; Motl & Schulte, 2015 ; Niet al., 2019 ; Zykov et al., 2022) . Of course retraining a new model from scratch to address each new predictive task of interest may be prohibitively laborious or expensive, which motivates a new class of foundation models (FMs) specif-  \n1University of Hong Kong, Shanghai X-Lab. Correspondence to: David Wipf \u003C[davidwipf@gmail.com](davidwipf@gmail.com) >.  \nProceedings of the 2 nd ICML Workshop on Foundation Models for Structured Data, Seoul, South Korea. 2026. Copyright 2026 by the author(s) .  \nically tailored for RDBs. Among others, these vary from LLM-based approaches that digest serialized RDB content (Wydmuch et al., 2024), to specialized Transformers pretrained based on in-context learning (ICL) (Fey et al., 2025 ; Hudovernik et al., 2026 ; Kothapalli et al., 2026), the common theme in each case being a frozen architecture capable of making predictions involving arbitrary unseen RDBs.  \nAs a robust and scalable alternative, it has also been demonstrated that certain parameter-free subgraph encoders combined with single-table foundation models can achieve near SOTA performance without requiring any RDB-specific pretraining (Xu et al., 2026) . And yet significant ambiguity still exists over RDB embedding and pre-training requirements for FMs. This is largely because of recent architectures with parameterized encoders that accept observable neighborhood labels as supplementary discriminative inputs (Chenet al., 2026 ; Hudovernik et al., 2026 ; Ranjan et al., 2026), a common technique from graph learning (Shi et al., 2020 ; Wang et al., 2021) . Performance results using such models point towards the continued efficacy of RDB-specific pre-training to optimally exploit these labels on new tasks.  \nTo help resolve this ambiguity, herein we closely examine notable factors influencing the performance of parameterized versus parameter-free RDB encoders, particularly when observable labels are available as encoder inputs. Our contributions are as follows:  \n• On the theoretical side, we analyze two complementary roles that labels may fill as encoder inputs, either as standard discriminative features or as a mechanism for determining the import","cbCaisPmQxKFCv7r","https://ap.wps.com/l/cbCaisPmQxKFCv7r","pdf",524167,1,15,"English","en",105,"# Abstract\n# Introduction\n# Basics of RDB Foundation Models\n## Predictive Modeling with Relational Context\n## Computing RDB Embeddings","[{\"question\":\"What problem does the document address for relational database (RDB) modeling?\",\"answer\":\"It studies how to predict missing or future values in a target column using heterogeneous tabular information stored in an RDB, ideally without retraining a new model for each new prediction task.\"},{\"question\":\"Why is the choice between parameter-free and parameterized RDB encoders still ambiguous?\",\"answer\":\"Recent work suggests parameter-free subgraph encoders with single-table foundation models can reach near SOTA, while other studies argue parameterized encoders pre-trained with observable neighborhood labels better learn task-specific representations.\"},{\"question\":\"What are the main theoretical and empirical contributions?\",\"answer\":\"Theoretical analysis shows limitations on how trainable encoder parameters can exploit label roles, and experiments demonstrate that much simpler parameter-free encoders still achieve strong performance across many relevant benchmarks.\"}]",1784177127,38,{"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":85,"head_meta":87,"extra_data":89,"updated_unix":27},"parameter-free-encoders-remain-viable-for-rdb-foundation-models","",{"@graph":35,"@context":84},[36,53,67],{"@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/parameter-free-encoders-remain-viable-for-rdb-foundation-models/81933/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does the document address for relational database (RDB) modeling?","Question",{"text":74,"@type":75},"It studies how to predict missing or future values in a target column using heterogeneous tabular information stored in an RDB, ideally without retraining a new model for each new prediction task.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"Why is the choice between parameter-free and parameterized RDB encoders still ambiguous?",{"text":79,"@type":75},"Recent work suggests parameter-free subgraph encoders with single-table foundation models can reach near SOTA, while other studies argue parameterized encoders pre-trained with observable neighborhood labels better learn task-specific representations.",{"name":81,"@type":72,"acceptedAnswer":82},"What are the main theoretical and empirical contributions?",{"text":83,"@type":75},"Theoretical analysis shows limitations on how trainable encoder parameters can exploit label roles, and experiments demonstrate that much simpler parameter-free encoders still achieve strong performance across many relevant benchmarks.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]