Lecheng Ke

Psychology + Data Science

Lecheng Ke.

Cognition & neuroscience

I’m interested in how brain activity shapes what we notice, remember, and decide—and why these processes differ across people.

Applying to PhD programs for Fall 2027. Current focus: my honors thesis on how large language models make causal judgments, with people as a comparison point.

I study at Emory University. My research experience spans behavioral experiments, emotional memory, visual perception, EEG processing, and computational modeling in ultrasound-related neuromodulation research. I’m also developing computational projects to explore the mechanisms behind these observations, including preliminary models of how external input and excitation–inhibition balance shape neural dynamics.

Research experience

Emory University Class of 2027 Atlanta

Live model

Two hundred neurons, running right now.

A 3D view of a spiking excitatory–inhibitory network based on Izhikevich (2003): 160 excitatory neurons and 40 inhibitory neurons arranged on a brain-like shape. Every flash is a spike, computed live in your browser. Drag to rotate.

Perturb the network yourself

Educational model written for this site. Not fitted to data.

A live 200-neuron spiking network runs here when JavaScript is available.

01 / Research interests

From behavior to mechanism.

01

Models &
neural representations

Computational neuroscience · NeuroAI

When a model predicts a response, what has it captured?

I want to understand when a computational or AI model captures something about how the brain represents information, rather than just producing a similar answer. Comparing models with behavior and neural data offers a way to ask where those similarities hold—and where they break down.

Experience behind the questionVisual-perception research in the Dilks Lab, resting-state EEG processing in TReAD, and preliminary neural-dynamics modeling.

02

Brain dynamics &
causal questions

Neural dynamics · Neuromodulation · Cognitive neuroscience

What could changing neural activity tell us about cognition?

I’m interested in how changes in a neural system affect perception, memory, or behavior, and more broadly how perturbing a neural circuit changes cognitive representations, learning, and decision-making. Brain stimulation interests me as a way to test causal explanations, while computational models can make predictions about how a change in input or excitation–inhibition balance might affect the system. I’m especially drawn to questions that separate timing from rate: for example, whether stimulation changes when a process succeeds rather than how often it succeeds.

Experience behind the questionAt GIIST, I explored preliminary BrainPy/Python simulations of how external input and excitation–inhibition balance shape brain-state dynamics, including parameter sensitivity and activity stability. My contribution described here is computational; using stimulation to test cognitive mechanisms is a direction I hope to pursue.

Try it

Raise the external drive, a stand-in for stimulation-like input, or weaken inhibition, and watch the same network move between sparse, asynchronous, and synchronized firing.

Educational model written for this site: the Izhikevich (2003) random network, with 160 excitatory and 40 inhibitory neurons and noisy thalamic input. Synaptic weights are scaled ×5 so each neuron receives the same mean input as in the original 1000-neuron network. Why ×5: the browser runs 200 neurons instead of the 1,000 in Izhikevich (2003), so each weight is multiplied by 1000/200 = 5, which keeps the total input each neuron receives from the network about the same. Not fitted to data.
03

Affect, motivation &
memory

Affective neuroscience · Individual differences

Why does the same experience affect people differently?

I’m interested in how emotion and motivation shape what people remember and the decisions they make. I want to understand how differences in reward, effort, and emotional memory might help explain differences in mood and motivation, including symptoms associated with depression.

Experience behind the questionBehavioral study support and EEG processing in TReAD, emotional-memory sessions and transcript coding in Hamann, and earlier work on cognitive effort in macaques.

04

Human cognition &
AI

Cognitive science · Language models · Causal reasoning

When a language model judges cause and effect, what is it relying on?

I’m interested in how large language models make causal judgments, and in what those judgments can and cannot tell us about reasoning. Causal judgment is a useful test case because a fluent, plausible answer can look like causal reasoning without being it. Research on how people judge cause and effect gives a point of comparison for asking what a model is actually doing.

Experience behind the questionMy honors thesis, on whether large language models judge a research finding as causal by the study design or by the way it is written.

02 / Research experience

Questions grounded in lab work.

TReAD Lab

Research assistant

Emory University

Supported studies of motivation, effort, and decision-making through participant screening, behavioral task administration, study coordination, and MRI visit support. Cleaned resting-state EEG data in EEGLAB, including filtering, channel review, and artifact review.

Hamann Lab

Research assistant

Emory University

Independently administered emotional-memory sessions and coded recall transcripts using established scoring schemes. Participated in reliability checks and maintained consistent participant-facing procedures.

Guangdong Institute of Intelligent Science and Technology (GIIST)

Research intern

Zhuhai, China

Explored preliminary BrainPy/Python simulations of how external input and excitation–inhibition balance might change neural dynamics, in the context of ultrasound-related neuromodulation research, including parameter sensitivity, activity stability, and network-level responses. Since the internship (May–August 2025), I have kept practicing this modeling remotely.

Dilks Lab

Research assistant

Emory University

Ran EyeLink setup, calibration, and gaze-data quality checks for work on blind-spot filling-in and short-term visual plasticity. Collected behavioral and eye-tracking data under standardized protocols, and contributed to a literature synthesis on GABA, adult V1 plasticity, and cortical reorganization for a broader fMRI/MRS study.

Comparative Primate Cognition Lab

Research assistant

Emory National Primate Research Center

Helped design controlled tasks asking whether rhesus macaques avoid working-memory effort when physical effort and reward are held constant. Contributed to proposals, task materials, and experimental sessions.

Griggs Lab

Research assistant

Emory University

Screened literature for a systematic review of brain health, cognition, and sleep; created REDCap instruments; and helped plan a public-data analysis. Dataset selection and the empirical analysis were not yet finalized.

Teaching, clinical exposure & service

Peer tutoring · Emory University

Individual and group tutoring in psychology, neuroscience, and quantitative coursework.

Clinical internship · Shenzhen Longhua People’s Hospital

Observed neurology and clinical psychology practice and supported coordination and participant engagement in a youth mental-health program.

Sprouting Readers & ELSA · Emory University

Mentored children through reading activities and supported campus library and wellness events.

Honors

Phi Eta Sigma National Honor Society · Dean’s List

03 / Computational projects

Models I’m learning from.

GitHub

Small working models and ongoing explorations, with their assumptions and current limits.

Interactive models load here when JavaScript is available. You can also read the neuron simulation source or open the brain-state prototype directly.
Honors thesisCausal judgments in people and LLMsIn progress

My honors thesis asks whether large language models judge a research finding as causal by the study design behind it or by the way the finding is written, with people’s judgments as a comparison point.

Human cognition & AI research

Where it stands

Past the pilot phase: six language models have answered questions about 1,000 research summaries, and the analysis is under way.

Computational learningNeuron dynamics in PythonWorking demo

A reproducible Izhikevich regular-spiking neuron simulation, implemented with NumPy and visualized with Matplotlib. It records membrane voltage, a recovery variable, and spike times under an applied current.

View repository

What I’m working toward

Exploring additional firing patterns and developing a BrainPy-native implementation. The current standalone demo uses NumPy; it is a learning model, not a fitted model of human neural data.

Independent concept projectBrain-state circuit resonanceSynthetic prototype

An exploration of state representations, memory cues, and feedback. The repository includes an interactive browser prototype and synthetic-data experiments that make assumptions about state similarity and uncertainty explicit.

Screenshot of the prototype: a dark 3D concept map of labelled nodes beside a control panel with cue-strength and feedback-gain sliders.Load the 3D prototype here View repository

Current limits

The simulations use synthetic data. They do not demonstrate human memory reconstruction, neural decoding performance, or clinical effects. I’m continuing to refine the modeling questions and document the assumptions.

Reading toolPreprint ReelsWorking prototype

A daily reading tool. Each morning an automated job picks the most relevant new arXiv paper, reads it in full, and writes a one- to two-minute vertical reel: the question, how it was tested, what was found, and the catch. It runs on Claude artifacts and scheduled tasks.

View repository
Reel hook scene: a paper title at the top and a large headline, Inside AI chatbots, researchers found a running best guess about hidden causes.Reel result scene: a large number, 0.83 to 0.99, with the line, A simple decoder could read the belief state from the activations.Reel end card: the full paper title and authors, a question to think about, and buttons for the abstract and PDF.
Hands-on learningPhysical AI explorationEarly exploration

An early hands-on project exploring what sensing, action, and interaction with the physical world add to intelligence.

04 / Coursework

Courses, term by term.

Psychology, data science, neuroscience, and related fields.

Enrollment as of September 3, 2026. Fall 2026 courses are in progress. This list covers courses taken at Emory; AP and test equivalencies are not included.

Fall 20265 coursesIn progress
  • DATASCI 495AHonors Research
  • DATASCI 530Computing I
  • ENVS 229Atmospheric Science with Lab
  • ENVS 485Special TopicsTopic: Ecosystems of the Southeast
  • PSYC 352The Genetics of Human Behavior
Spring 20266 coursesCompleted
  • DATASCI 220Regression Analysis
  • DATASCI 350Data Science Computing
  • ENVS 385Topics: Environmental StudiesTopic: Restoration Ecology
  • JPN 102Elementary Japanese II
  • NBB 318Introduction to Feminist Neuroscience
  • PSYC 499RDirected Research
Fall 20257 coursesCompleted
  • ENVS 131Introduction to Environmental Studies Field Studies
  • ENVS 232Fundamentals of Ecology with Lab
  • JPN 101Elementary Japanese I
  • PSYC 499RDirected Research
  • QTM 210Probability and Statistics
  • QTM 340Approaches to Data Science with Text
  • QTM 385Special Topics: QTMTopic: Experimental Methods
Spring 20256 coursesCompleted
  • PSYC 200WLab in Experimental Methods
  • PSYC 324Sleep & Dreaming, Brain & Mind
  • PSYC 353Behavioral Neuroscience
  • PSYC 499RDirected Research
  • PSYC 770RTopic Seminars: PsychologyTopic: Cortical Plasticity
  • QTM 151Introduction to Statistical Computing II
Fall 20246 coursesCompleted
  • BIOL 141Foundations of Modern Biology I
  • BIOL 141LFoundations of Modern Biology I Lab
  • MATH 210Advanced Calculus for Data Sciences
  • NBB 201Foundations of Behavior
  • PSYC 207Brain & Behavior
  • QTM 110Introduction to Scientific Methods
Spring 20246 coursesCompleted
  • ENVS 140Environmental Change & Health
  • HIST 190Freshman SeminarTopic: Fake News 2015–Present
  • PE 167Beginning Yoga
  • PSYC 111Introduction to Psychology II
  • PSYC 209Perception and Action
  • QTM 150Introduction to Statistical Computing I
Fall 20237 coursesCompleted
  • CHN 271Modern China in Film & Fiction
  • ECS 101Emory Edge
  • HLTH 100It’s Your Health & Wellbeing
  • MATH 221Linear Algebra
  • PSYC 110Introduction to Psychology I: Psychobiology & Cognition
  • PSYC 212Social Psychology
  • QTM 100Introduction to Statistical Inference

05 / Contact

Let’s talk about
cognition and the brain.

I’m applying to PhD programs for Fall 2027.

rola.ke@emory.eduGitHub projectsRequest my CV