Summary of Evaluating Ai-generated Essays with Gre Analytical Writing Assessment, by Yang Zhong et al.
Evaluating AI-Generated Essays with GRE Analytical Writing Assessment
by Yang Zhong, Jiangang Hao, Michael Fauss, Chen Li, Yuan Wang
First submitted to arxiv on: 22 Oct 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 The recent breakthrough in generative AI enables large language models (LLMs) to produce realistic and coherent texts. While various metrics assess text quality, a lack of rigorous evaluation persists for complex writing assessments. This study examines essays generated by ten leading LLMs for the Graduate Record Exam’s analytical writing assessment. Essays were evaluated using human raters and the e-rater automated scoring engine. Top-performing models Gemini and GPT-4o received average scores of 4.78 and 4.67, respectively, falling between “generally thoughtful” and “competent analysis.” The study also evaluated detection accuracy of these essays. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper studies how good large language models are at writing like humans do. They use special machines to grade the writing, just like they do in school tests. They looked at 10 different AI models and saw which one did the best job. Two models, Gemini and GPT-4o, were really good and got high scores on the test. The study also tried to figure out if these AI-written essays can be detected as being written by machines. |
Keywords
» Artificial intelligence » Gemini » Gpt
