Evaluating German Transformer Language Models with Syntactic Agreement Tests

TL;DR

This paper evaluates German transformer language models using syntactic agreement tests, revealing their strengths and limitations in handling complex German syntactic structures.

Contribution

It introduces German-specific agreement tasks and analyzes the performance of state-of-the-art German TLMs on these tasks.

Findings

German TLMs perform well on agreement tasks

Certain complex syntactic structures challenge TLMs

Analysis highlights areas for future improvement

Abstract

Pre-trained transformer language models (TLMs) have recently refashioned natural language processing (NLP): Most state-of-the-art NLP models now operate on top of TLMs to benefit from contextualization and knowledge induction. To explain their success, the scientific community conducted numerous analyses. Besides other methods, syntactic agreement tests were utilized to analyse TLMs. Most of the studies were conducted for the English language, however. In this work, we analyse German TLMs. To this end, we design numerous agreement tasks, some of which consider peculiarities of the German language. Our experimental results show that state-of-the-art German TLMs generally perform well on agreement tasks, but we also identify and discuss syntactic structures that push them to their limits.