Evaluating Large Language Models: Challenges and Methods

• Room 119B, Philadelphia Convention Center, PA, USA
• 2-6pm, February 25, 2025

Presenters

• Jindong Wang, William & Mary, jwang80 (at) wm.edu
• Kaijie Zhu, UC Santa Barbra, kaijiezhu (at) ucsb.edu
• Linyi Yang, Westlake University, yanglinyi (at) westlake.edu.cn
• Yue Feng, University of Birmingham, yueyuef (at) outlook.com
• Yue Zhang, Westlake University, zhangyue (at) westlake.edu.cn

Contents at a glance

Slides are available to download: Google Drive

Partial video recording

• Background and challenges
• Evaluation approaches
  • Adversarial and OOD robustness
  • Dynamic evaluation
  • Reasoning
  • Safety
  • User evaluation
• Evaluation in social science
• Conclusion

Acknowledgement