Science progresses by building upon the prior body of knowledge documented in scientific publications. The acceleration of research makes it hard to stay up-to-date with the recent developments and to summarize the ever-growing body of prior work. To address this, the task of citation text generation aims to produce accurate textual summaries given a set of papers-to-cite and the citing paper context. Due to otherwise rare explicit anchoring of cited documents in the citing paper, citation text generation provides an excellent opportunity to study how humans aggregate and synthesize textual knowledge from sources. Yet, existing studies are based upon widely diverging task definitions, which makes it hard to study this task systematically. To address this challenge, we propose CiteBench: a benchmark for citation text generation that unifies multiple diverse datasets and enables standardized evaluation of citation text generation models across task designs and domains. Using the new benchmark, we investigate the performance of multiple strong baselines, test their transferability between the datasets, and deliver new insights into the task definition and evaluation to guide future research in citation text generation. We make the code for CiteBench publicly available at https://github.com/UKPLab/citebench.

This paper describes a method for retrieving evidence and predicting the veracity of factual claims, on the FEVEROUS dataset. The evidence consists of both sentences and table cells. The proposed method is part of the FEVER shared task. It uses similarity scores between TF-IDF vectors to retrieve the textual evidence and similarity scores between dense vectors created by fine-tuned TaPaS models for tabular evidence retrieval. The evidence is passed through a dense neural network to produce a veracity label. The FEVEROUS score for the proposed system is 0.126.