Haotian Qi [齊]

h.t.qi[at]outlook.com

Hello! My name is Haotian Qi. I completed my Bachelor's and Master's studies at the University of Sheffield, where I received First Honors for my Master of Computing with AI and CS degree. My focus is on NLP and LLM

I am interested in researching the complex reasoning abilities of Large Language Models (LLMs). Specifically, I want to better understand the core reasons behind their abilities and explore ways to enable complex and accurate reasoning for smaller models like LLaMA. I am curious about how we can achieve GPT-4's capacity with minimal cost in smaller scale LLMs. Additionally, I am keen to explore the use of multimodal input to enhance conversation models' interaction with LLMs. By using LLMs to receive multiple types of input such as speech, human action, or facial expressions, we can create more natural and continuous responses with AI through conversation.

...

Lingustic Information Within Large Language Models
NLP BERT Probing Linguistic Features

Positional Excitation Implementation for Human Action Recognition with ACTION-Net Model
Computer Vision Self-attention Action Recognition

Research Interest

  • Efficient LLMs
  • Reasoning in LLMs
  • Multimode
  • Conversation Model