Hi! I am Huadong

What is intelligence? One way to understand it is through the lens of learning. Humans observe sparse and ambiguous data, yet from those, humans develop rich internal models to make inferences about the world. Given just a few in-context examples, large language models (LLMs) can also generalize to solve novel tasks. I study the behavior of humans and LLMs to understand their inductive biases and how those guide their learning. What makes a good inductive bias for guiding learning, and how do inductive biases constrain learning, leading to good or bad generalization?

Through this framework, science can also be viewed as a learning process, one that constructs priors for humans and AI to learn from. What is the best form of knowledge that explains the world but also fits the structure of the human mind? How can we describe the space of all possible forms of knowledge, constrained by the structure of the human mind?

I am a postdoc with James Evans at the University of Chicago, studying these questions in the context of AI for science and AI alignment.

During my PhD in Psychology at the Georgia Institute of Technology, advised by Robert Wilson, I asked a similar question about the concepts researchers use to explain human decisions. I used recurrent neural networks, with fewer assumptions about decision making, to examine what conventional models treat as noise and revise the vocabulary those models use. This led to an online learning perspective: people learn both about the world and how to change their learning strategies, with strategy changes resembling policy gradient ascent.

I also studied in-context learning in LLMs, where a model learns from a prompt without changing its trained weights. This let me examine how existing representations shape new learning and cause interference between memories. My PhD collaborations included Marcelo Mattar at NYU and Li Ji-An at UCSD, with part of my LLM work done with Xue-Xin Wei at UT Austin and Kwonjoon Lee at Honda Research Institute. More details are on my Projects page.

Before my PhD, I worked with Da-Hui Wang at Beijing Normal University, developing spiking neural network models to investigate the biophysical mechanisms of binding features across two dimensions in multi-item working memory. With Xue-Xin Wei, I developed a recurrent neural network model to investigate the neural mechanisms underlying how working memory represents prior distributions over stimuli. Earlier, I used EEG to study working memory and studied counseling psychology with a focus on cognitive behavioral therapy.

The pronunciation of my name is “HWAH-doang SHAWNG”.

I am on the job market

I am looking for full-time Research Scientist and faculty positions in using AIs to model human behavior (including the science process) and AI alignment.

Want to collaborate or chat?

I’m always happy to collaborate. Whether you’re interested in my research or just want to chat about science, shoot me an email. If you’re at UChicago, even better. I’m usually at the Knowledge Lab (1155 E 60th St), Room 203. Feel free to just walk in.

Some of my writing

I also write about cognition, science, and the concepts we use to understand the world. If something interests you, feel free to email me. I’d like to hear what you think.

Some interesting facts:

• I enjoyed reading when I was young. My favorite writers are James Joyce, Milan Kundera, Jorge Borges, Franz Kafka, Dostoyevsky and Edgar Allan Poe.
• I chose this stupid username (sakimarquis) as a teenager. It came from Saki and Márquez.
• Yet I haven’t read much since high school. I suddenly lost my patience with long books.
• I am addicted to computer games, but only when an exam is approaching. Since I have few exams to take, I seldom play them now.
• I enjoy sick jokes and embarrassing short videos.
• Using a second language is painful for me, mostly because I can’t help being sarcastic but I can’t do it well in English.
• I am bad at calculating. I often mess up single-digit calculations, even with a pen and paper. This always makes me doubt myself as a researcher in computational neuroscience. (But large language models also fail at simple calculations, so I am not alone.)