In this work, we address question answering (QA) over a hybrid of tabular and textual data
that are very common content on the Web (eg SEC filings), where discrete reasoning …

Question answering over hybrid contexts is a complex task, which requires the combination
of information extracted from unstructured texts and structured tables in various ways …

Question answering requiring numerical reasoning, which generally involves symbolic
operations such as sorting, counting, and addition, is a challenging task. To address such a …

Discrete reasoning over table-text documents (eg, financial reports) gains increasing
attention in recent two years. Existing works mostly simplify this challenge by manually …

Answering questions about hybrid data combining tables and text is challenging. Recent
research has employed encoder-tree decoder frameworks to simulate the reasoning …

Large Language models (LLMs) usually rely on extensive training datasets. In the financial
domain, creating numerical reasoning datasets that include a mix of tables and long text …