Summary of Ten Words Only Still Help: Improving Black-box Ai-generated Text Detection Via Proxy-guided Efficient Re-sampling, by Yuhui Shi et al.
Ten Words Only Still Help: Improving Black-Box AI-Generated Text Detection via Proxy-Guided Efficient Re-Sampling
by Yuhui Shi, Qiang Sheng, Juan Cao, Hao Mi, Beizhe Hu, Danding Wang
First submitted to arxiv on: 14 Feb 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
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 method estimates word generation probabilities as pseudo white-box features via multiple re-sampling to improve AI-generated text (AIGT) detection in the black-box setting. The POGER (Proxy-Guided Efficient Re-Sampling) approach selects a small subset of representative words and performs multiple re-sampling to enhance AIGT detection. This method outperforms existing baselines on various datasets, including texts from humans and seven large language models, in terms of macro F1 score while maintaining lower re-sampling costs. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps solve problems like fake news and academic dishonesty by improving AI-generated text detection. It shows a new way to detect if text was written by a human or a computer. The method works well even when the AI model’s internal workings are not known. This is important because many AI-generated texts look very similar to those written by humans. |
Keywords
* Artificial intelligence * F1 score
