Mingmeng Geng - Analysis and Detection of LLM-Generated Content
- Date: 09 février 2026 à 13h
- Salle: 65-66 304
With the rapid advancement and widespread adoption of large language models (LLMs), the volume of content they generate is growing. Therefore, understanding the characteristics of such content has become increasingly important, and many researchers are attempting to develop corresponding detection methods. In this presentation, I will focus on the analysis and detection of LLM-generated content in academia. Although their traces can be found from multiple perspectives, accurate detection may be an infeasible task in many cases. However, if we shift the perspective from detection to estimating the impact of LLMs, what was once impossible may become possible.