Feedback Generation

Feedback generation, the automated creation of instructive critiques for various tasks, aims to improve learning and efficiency across diverse fields. Current research heavily utilizes large language models (LLMs), often employing techniques like retrieval augmented generation (RAG) and fine-tuning on synthetic or real-world datasets to enhance feedback quality, specificity, and alignment with pedagogical goals. This work spans applications from educational assessment (e.g., grading essays and programming assignments) to scientific writing and even movement analysis in rehabilitation, with a focus on optimizing feedback for clarity, actionability, and learner-specific needs. The ultimate goal is to create effective, scalable, and personalized feedback systems that augment human capabilities in teaching, learning, and other knowledge-intensive tasks.