Department of Psychological and Brain Sciences,
Johns Hopkins University,
Baltimore MD, 21218
Research Trajectory
Hey, this is Xian! I’m a Ph.D. candidate at the Department of Psychological & Brain Sciences, Johns Hopkins University. I am co-advised by Dr. Janice Chen and Dr. Christopher Honey. My research leverages naturalistic paradigms to investigate how we process ongoing events and how these experiences are encoded and retrieved using behavioral measures, computational analysis, and fMRI.
Primary Projects
Humans are agents: our choices actively shape the trajectory of events in our lives. These choices rely on personal knowledge and preferences that can affect later memory. In this behavioral project, we studied how people’s memory of a naturalistic sequence of events is altered when their choices control the future using choose-your-own-adventure stories. We also examined how memory performance is affected in the face of denial of choices.
We are collecting fMRI data of subjects doing the choose-your-own-adventure stories. This follow-up study aims to understand the neural associations for agency's effects on memory.
Our thoughts are continuous on the scale of seconds and minutes, despite internal and external distractions. How are we able to preserve our own mental contexts in the face of such common interruptions? In this fMRI study, participants listened to narratives (intact or scrambled) that were interrupted (or not) by silent pauses or theory-of-mind tasks.
Results suggest that aisde from hippocampal encoding, our brains could possess a top-level context representation acting like an "episodic background", which is continually and automatically updated in the background and is interference-resistant, likely via representation reversal. Collectively, results suggest two pathways to maintaining mental continuity, one through standard hippocampal encoding and reload, and another via "episodic background".
Studies of scene perception often focus on brief visual presentation (<1 second). But it may take several seconds, or even minutes, to make sense of a complex scene in real life. As each person takes their own “exploration path” through a visual scene, how does the history of their path influence their understanding of the next moment? In this fMRI study, we tested whether exploration paths during the perception of complex images could predict neural responses within and across individuals.
Other Projects
AI rate causal relations: ChatGPT agent does causal rating for narrative events on par with humans
Try out the Causal Rater for Story on an event-segmented story. This also works with movie descriptions. See an example input below.
Thinking out-loud (fMRI): Neural dynamics of spontaneous memory recall and future thinking in the continuous flow of thoughts
The human brain constantly recalls past events and anticipates future ones, generating a continuous flow of thoughts. But what neural mechanisms underlie this process? In this fMRI project, we examined the neural dynamics associated with the natural transitions and flow structure of spontaneous memory and future thinking. Find out more below.
Contact Me
To learn about the CYOA paradigm and how you could incorporate agency in your research study.
To discuss developing ChatGPT agent for rating tasks in naturalistic experiments that can perform on par with human participants.
To discuss the neural mechanisms and computations underlaying naturalistic events comprehension and memory encoding/retrieval.
To chat about how our work could relate to each other!
Find out more related naturalistic work on human cognitive abilities @Chen lab.
Find out more related memory work using neural and computatioanl methods @Honey lab.
@ 2025, Xian Li
xianl.cogneuro@gmail.com