Haoran Yang
Post-graduate student @ The Chinese University of Hong Kong
generalization
large language models
llms
rephrasing
prompt tuning
in-context learning
other foundations of machine learning
learning & optimization for snlp
interpretability & analysis of nlp models
other foundations of speech & natural language processing
commonsense generation
order issue
compositional instruction
memory-efficient fine-tuning
fine-tuning
7
presentations
5
number of views
SHORT BIO
Haoran Yang is a candidate postgraduate student at the Chinese Univerisity of Hong Kong, supervised by Professor Wai Lam. I am affiliated with the CUHK Text Mining Group. My research mainly focuses on Natural Language Processing (text generation, pretrained language models, parameter-efficient tuning).
Presentations
Chain-of-Dictionary Prompting Elicits Translation in Large Language Models
Hongyuan Lu and 5 other authors
A Thorough Examination of Decoding Methods in the Era of LLMs
Chufan Shi and 6 other authors
Unveiling the Generalization Power of Fine-Tuned Large Language Models
Haoran Yang and 5 other authors
Rephrasing Invokes Better Generations for Large Language Models
Haoran Yang and 2 other authors
Exploring Compositional Generalization of Large Language Models
Haoran Yang and 3 other authors
Bridging the Gap between Pre-Training and Fine-Tuning for Commonsense Generation
Haoran Yang
On the Effectiveness of Parameter-Efficient Fine-Tuning
Zihao Fu and 5 other authors