arXiv：cs.AI· Zheyu Zhang, Gjergji Kasneci·· 22 小时前
TabWM：将关系学习重新定义为世界建模
Relational Learning as World Modeling
AI 导读
研究者提出 TabWM（Tabular World Model）与 WorldRel 基准，让模型将未见系统的执行记录作为上下文，无需更新参数即可预测新动作引发的稀疏结构化编辑。在 512 个留出世界上，完整编辑 F1 从无执行记录时的 0.016 升至 64 条记录时的 0.685。由于预测结果是显式关系编辑，可递归应用以想象候选未来并在行动前比较动作，更广的搜索提升了所选动作序列的实际回报。
正文
Abstract:Relational learning typically predicts from a database state, but decision making asks a different question: how will that state change under an action? In operational systems, the answer depends on system-specific dynamics that may not be visible in the database itself. We view each such system as a relational "world" and ask whether a model can infer how it behaves from observed executions. We introduce TabWM (Tabular World Model) and WorldRel, a generator of executable relational worlds for pretraining. Given executions from an unseen world, TabWM uses them as context to predict the sparse structured edits caused by a new action without parameter updates. Across 512 held-out worlds, complete-edit F1 rises from 0.016 with no observed executions to 0.685 with 64, where a correct edit must match its operation, relational location, and value. Because these predictions are explicit relational edits, they can be applied recursively to imagine candidate futures and compare actions before they are taken. Broader search over these predicted futures improves the realized return of selected action sequences. Together, these results suggest that relational dynamics can be learned in context, broadening relational learning from modeling system state to modeling how that state changes under action.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.16763 [cs.LG] |
| (or arXiv:2610.16763v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.16763 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zheyu Zhang [view email]
[v1]
Mon, 12 Oct 2026 13:24:37 UTC (555 KB)
来源：arXiv：cs.AI · arxiv.org