Tag
论文提出 TabWM（表格世界模型）与 WorldRel 生成器，通过在上下文中观察执行样本来推断关系型系统的动态，预测动作导致的稀疏结构化编辑，从而将关系学习从状态建模扩展到世界建模与动作规划。
This paper evaluates relational foundation models on high-cardinality data, revealing that context window limitations lead to performance drops, which can be mitigated by simple pre-aggregation steps.
The paper investigates injecting frozen relational transformer embeddings into an LLM via soft tokens, reporting negative results due to performance instability and sensitivity to serialization formats.
The paper proposes a Relational Hypergraph Transformer (RHT) architecture for complex multi-table analysis in healthcare, addressing five dimensions of complexity with a unified approach and sparse attention mechanism. It includes formal analysis, open-source implementation, and empirical evaluation on synthetic electronic health records.
Prior Labs releases open-source tools including RelArena-α, TabPFN-Rel, and RPI to advance reproducible relational learning, providing a unified benchmark framework and a TabPFN-based model.
This paper introduces relational structural causal models, extending structural causal models to settings with varying objects and relations. It provides theoretical results for identification and proposes relational neural causal models that outperform non-relational baselines on simulated traffic scenes.
This paper proposes a method for improving reasoning in large language models by recoding data to explicitly represent relationships, enabling efficient principled reasoning with polynomial-time learnability for relational rules, which addresses hallucinations and supports sound reasoning across multiple calls.