Summary of Large Language Models As Moocs Graders, by Shahriar Golchin et al.
Large Language Models As MOOCs Graders
by Shahriar Golchin, Nikhil Garuda, Christopher Impey, Matthew Wenger
First submitted to arxiv on: 6 Feb 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper investigates the feasibility of using large language models (LLMs) to replace peer grading in massive open online courses (MOOCs). The researchers focus on two state-of-the-art LLMs, GPT-4 and GPT-3.5, across three distinct MOOCs: Introductory Astronomy, Astrobiology, and the History and Philosophy of Astronomy. To instruct LLMs, they use a variant of the zero-shot chain-of-thought (Zero-shot-CoT) prompting technique with different prompts based on instructor-provided answers and rubrics. The results show that Zero-shot-CoT, when integrated with instructor-provided answers and rubrics, produces grades more aligned with those assigned by instructors compared to peer grading. However, the History and Philosophy of Astronomy course proves to be more challenging in terms of grading. This study suggests a promising direction for automating grading systems for MOOCs, especially in subjects with well-defined rubrics. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how computers can help grade homework assignments in online courses that have many students. Right now, teachers rely on other students to grade each other’s work, but this method isn’t very reliable. The researchers tested two advanced computer programs (GPT-4 and GPT-3.5) to see if they could do a better job of grading. They used these computers to grade homework assignments in three different courses: astronomy, biology, and the history of astronomy. The results show that when teachers provide answers and guidelines for the computers, they can produce grades that are very close to what human teachers would give. This study suggests that using computers to grade homework could be a good way to make grading more accurate and fair. |
Keywords
* Artificial intelligence * Gpt * Prompting * Zero shot
