Summary of Spotting Ai’s Touch: Identifying Llm-paraphrased Spans in Text, by Yafu Li et al.
Spotting AI’s Touch: Identifying LLM-Paraphrased Spans in Text
by Yafu Li, Zhilin Wang, Leyang Cui, Wei Bi, Shuming Shi, Yue Zhang
First submitted to arxiv on: 21 May 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 proposed novel detection framework, Paraphrased Text Span Detection (PTD), aims to identify paraphrased text spans within a text by assigning each sentence with a score indicating the paraphrasing degree. This approach differs from traditional text-level detection methods. To evaluate PTD models, a dedicated dataset called PASTED was constructed for paraphrased text span detection. Results demonstrate the effectiveness of PTD models in identifying AI-paraphrased text spans, both in-distribution and out-of-distribution. Statistical analysis highlights the crucial role of surrounding context in determining paraphrasing degrees. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about a new way to detect when someone has rewritten some text using artificial intelligence (AI). It’s like a plagiarism detector, but it looks for AI-generated text instead. The researchers created a special dataset called PASTED to test their method and found that it works well both when the AI is generating similar text and when it’s generating very different text. |
