Large language models are increasingly used to generate writing feedback, but according to a new arXiv paper, there has been no systematic comparison of LLM and expert feedback on the dimensions that writing research identifies as central. The paper, titled FOXGLOVE, sets out to address this gap by focusing on argumentative essays.
The study's title points to two key aspects of feedback: goal-oriented and anchored. Goal-oriented feedback likely refers to guidance tied to specific writing objectives, while anchored feedback may be linked to particular parts of the text. The abstract indicates that the work aims to understand how expert and LLM feedback compare along these dimensions.
Because the abstract is brief, the specific findings are not detailed here. However, the paper's motivation is clear: to provide a systematic comparison that has been missing from the literature. This could help clarify where LLM feedback aligns with expert practice and where it falls short, which is relevant for educational tools and writing instruction.