H2Table introduces a hierarchical hypergraph representation for complex tables, enabling a hypergraph encoder to capture semantic relationships between headers and cells. The framework uses learnable query vectors to extract structural embeddings for large language models. Experiments on the HiTab dataset show a 22.88% improvement over state‑of‑the‑art baselines on tables with four levels of nesting.
Tables are ubiquitous across diverse domains, yet reasoning over them remains a significant challenge for modern large language models (LLMs). Current approaches typically linearize tables into sequen...
arXiv:2606. 28916v1 Announce Type: cross Abstract: We introduce GRAB, a constructor-encoder-bridge pipeline for table question answering.
PARTAB is a framework that improves large language model reasoning on tables by constructing a structured evidence interface. It represents query‑relevant evidence as semantically coherent, row‑linked table regions and performs hierarchical selection over column groups and row‑level partitions before composing the evidence for answer generation. Evaluations on multiple table reasoning benchmarks show that PARTAB consistently outperforms full‑table prompting and recent methods, achieving strong performance on WikiTableQuestions and TabFact while remaining competitive on numerical reasoning tasks.
The paper introduces Table Graph Reasoner (TabGR), a model that represents tables as an Attributed Table Graph (ATG) to preserve row-column-cell structure and enable graph-based reasoning without task-specific training. It also proposes a Question-Guided Personalized PageRank (QG-PPR) mechanism to rerank tabular data and address the lost-in-the-middle issue. Experiments on multiple table reasoning benchmarks show that TabGR outperforms state-of-the-art models by up to 9.7% in accuracy.
arXiv:2506. 18421v3 Announce Type: replace-cross Abstract: The majority of data in businesses and industries is stored in tables, databases, and data warehouses.
Large Language Models (LLMs) have shown strong capabilities in table reasoning, but their effectiveness degrades as tables grow in size and complexity due to irrelevant context and difficulty localizi...
arXiv:2609.26658v1 Announce Type: cross Abstract: Integrating heterogeneous datasets within data lakes is a critical challenge, particularly for semantically related tables that lack the explicit att...
arXiv:2605. 20254v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have shown promising results on NLP tasks, however, their performance on tabular data still needs research attention, because Table Question-Answering (TQA) requires precise cell retrieval and multi-step structured reasoning.
MetaRTL is a two-stage framework for relational table learning that first generates lightweight pre-trained table embeddings and then applies non‑parametric message passing to extract meta‑path features. These features are aggregated using an attention module called MetaAttn, shifting computation from deep GNN stacks to efficient meta‑path aggregation. Experiments on 10 real‑world datasets across 24 tasks show that MetaRTL achieves strong performance while reducing computational cost.
arXiv:2610.34866v1 Announce Type: new Abstract: Large Language Models (LLMs) have shown remarkable capabilities across tasks, but they still struggle with complex reasoning that requires integrating...
arXiv:2603.28773v2 Announce Type: replace-cross Abstract: Large language models (LLMs) frequently generate confident yet factually incorrect content when used for language generation (a phenomenon of...
arXiv:2604. 12503v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) have shown remarkable capabilities across various tasks but remain prone to hallucinations in knowledge-intensive scenarios.