Abstract: In this article, we present BenchING, a new benchmark for evaluating large language models (LLMs) on their ability to follow structured output format instructions in text-based procedural ...
Switching from Google Docs to Zettlr finally gave me control over drafts, versions, and long-form writing.