Simon Yu

Simon Yu

MRes Student at NLP and AI

University of Edinburgh

Biography

Hi! I’m Simon Yu, a MRes student at the ILCC at University of Edinburgh, supervised by Dr. Jeff Z. Pan. I am also fortuantely work closely with Dr. Pasquale Minervini. I am broadly interested in NLP, generalisation in NLP, and robust and trustworthy AI. My bachelor’s degree was fully funded by the Macao Foundation Undergraduate Scholarship, (top 1% of students in A-level).

Sometimes we have plans, but sometimes life has other plans. For so long I’ve wished and obsessed upon certain opportunities, just to get other ones where in hindsight. “Trust the process” is so much deeper than what people think. Learning to let go is one of the most powerful research lessons.

Acknowledgement: Since I began my research, I have met many intelligent, disciplined, and wonderful peers to work with, including (but not limited to) Andrej Jovanovic@Cambridge, Jie He@Edinburgh, Hanxu Hu@Edinburgh, Chenmien Tan@Edinburgh, Yijun Yang@Edinburgh, Sanjukta Bhattacharya@Edinburgh, and Liangyu Chen@NTU. I have truly learned a lot from them, and I enjoyed all the discussion we had.

Interests
  • Building Trustworthy LLM
  • Instruction Tuning for Generlization
  • Representation Learning (Gradient)
Education
  • Master by Research in NLP, 2023-2024

    The University of Edinburgh

  • BSc in Artificial Intelligence and Computer Science, 2019-2023

    The University of Edinburgh

Experience

 
 
 
 
 
Research Assistant
October 2022 – September 2023 Edinburgh, United Kingdom
 
 
 
 
 
Research Intern
Noah Arks Lab R&D, Edinburgh
March 2022 – September 2022 Edinburgh, United Kingdom

Publications

(2024). Evaluating the Adversarial Robustness of Retrieval-Based In-Context Learning for Large Language Models. Submitted to COLM, Under Review.

(2024). kNN-XICL: Improving Cross-lingual In-Context Learning Calibration with Nearest Neighbor Inference. Committed to ACL 2024.

(2023). HJCL: Hierarchy-aware Joint Supervised Contrastive Learning for Hierarchical Multi-Label Text Classification. In EMNLP 2023.

PDF Cite Code

(2023). BUCA: A Binary Classification Approach to Unsupervised Commonsense Question Answering. In ACL 2023.

PDF Cite Slides