# Time Travel in LLMs: Tracing Data Contamination in Large Language Models

**Authors:** Shahriar Golchin, Mihai Surdeanu  
**arXiv:** [2308.08493](https://arxiv.org/abs/2308.08493)  
**Date:** 2023-08-16  
**Summary page:** https://www.summarizepaper.com/en/paper/2308.08493v1/

## Key points

- Researchers propose method to identify data contamination in LLMs through guided instruction prompts and performance metrics
- Approach involves assessing individual instances and entire dataset partitions for potential contamination
- GPT-4 is evaluated using few-shot In-context Learning Prompt and compared to reference text
- Detailed descriptions of datasets like IMDB Movie Reviews, AG News, WNLI, SAMSum, and XSum provide context for the study
- Findings reveal contamination in GPT-4 with AG News, WNLI, and XSum datasets
- High accuracy achieved in detecting contamination compared to manual evaluation by human experts

## Summary

Researchers propose method to identify data contamination in LLMs through guided instruction prompts and performance metrics. The approach involves assessing individual instances and entire dataset partitions for potential contamination. GPT-4 is evaluated using few-shot In-context Learning Prompt and compared to reference text. Detailed descriptions of datasets like IMDB Movie Reviews, AG News, WNLI, SAMSum, and XSum provide context for the study. Findings reveal contamination in GPT-4 with AG News, WNLI, and XSum datasets. High accuracy achieved in detecting contamination compared to manual evaluation by human experts.