About Zhiheng
Hi, here is Zhiheng. Welcome to my personal website, nice to see you😆~ I’m currently seeking a PhD position on the term of 24Fall. My research interests are specifically focused on structural text planning and generation, causality in natural language processing, and the development of efficient large language models (LLMs). I am deeply passionate about advancing the field of AI by creating innovative systems that merge language structures with logical reasoning to address complex real-world problems efficiently. – for more details, please see my SoP and CV.
Currently I’m a senior undergraduate student major in Computer Science in University of Hong Kong, supervised by Lingpeng Kong. Last summer, I work at the Berkeley NLP, supervised by Kevin Yang and Prof Dan Klein, focusing on fact tracking and contradict detection in a narrative structure, such as story outline. Previously, I’ve worked at the ETH NLPED Lab under the supervision of Zhijing Jin and Dr. Mrinmaya Sachan, and as a research intern at Megvii.
Research Highlight
- Fact tracking and contradict detection in story outline
- Probing the Understanding of Large Language Model in Causality
- Explore new Training Paradigm for a Causal Graph based on Conditional Independent Test
- Construct and Pre-process Datasets on NLP for Social Good
Publications
Can Large Language Models Infer Causation from Correlation?
Published in , 2023
This research introduces the first benchmark dataset, Corr2Cause, to test large language models (LLMs) pure causal inference skills.
Recommended citation: Jin Z, Liu J, Lyu Z, et al. Can Large Language Models Infer Causation from Correlation? arXiv preprint arXiv:2306.05836, 2023. https://arxiv.org/abs/2306.05836
Can Large Language Models Distinguish Cause from Effect?
Published in , 2023
Our paper conducts a post-hoc analysis to check whether large language models can be used to distinguish cause from effect.
Recommended citation: Jin, Lalwani, A., Vaidhya, T., Shen, X., Ding, Y., Lyu, Z., Sachan, M., Mihalcea, R., & Schölkopf, B. (2022). Logical Fallacy Detection. https://openreview.net/forum?id=ucHh-ytUkOH
Psychologically-Inspired Causal Prompts.
Published in , 2023
This paper is about a prompting method embedded causal direction and analyze the performance gap of LLMs
Recommended citation: Lyu, Z., Jin, Z., Mattern, J., Mihalcea, R., Sachan, M., & Schoelkopf, B. (2023). Psychologically-Inspired Causal Prompts. arXiv preprint arXiv:2305.01764. https://arxiv.org/pdf/2305.01764
Logical Fallacy Detection
Published in , 2022
This paper is about the a dataset of Logical Fallacy Detection and its baseline model
Recommended citation: Jin, Lalwani, A., Vaidhya, T., Shen, X., Ding, Y., Lyu, Z., Sachan, M., Mihalcea, R., & Schölkopf, B. (2022). Logical Fallacy Detection. https://arxiv.org/abs/2202.13758
More
If you’re a potential collaborator with similar interests, feel free to email me at zhihenglyu.cs@gmail.com. I’m open to discussing my current projects or exploring potential cooperation in any of the following areas: Research Project, Developing Open source project, Cofounder of Community/Association, Start-up Companies.
For freshmen and sophomores interested in pursuing CS research, please don’t hesitate to reach out. I’m excited to meet new people from diverse backgrounds, share my experiences, and potentially start a mentorship.