Research in Computational Cognition
- Faculty Sponsor: Hongjing Lu
- Department: Psychology
- Contact Name: Hongjing Lu
- E-mail: hongjing@ucla.edu
Description of Research Project
We study human perception and reasoning from a computational perspective. We are especially interested in what prior knowledge humans assume in making an inference from few examples. Our research aims to develop computational models and test AI models for a range of perception and reasoning experiments, and assess the validity of computational models by comparing their predictions with human performance in controlled experiments.
Our current areas of active study include action understanding, object recognition, causal reasoning, and analogical reasoning. See check our lab website https://cvl.psych.ucla.edu/publications/ for papers related to recent projects.
Description of Student Responsibilities
Students selected for this project should have a strong interest in computational cognition, with interests spanning both human cognition and AI, as well as strong programming skills. They should be familiar with implementing AI models, including using APIs for commercial AI models and Python programming for open-source models. A strong ability to read and understand research papers with mathematical formulations is also expected. Students are expected to have strong organizational skills and enjoy working collaboratively on research projects.
Students will be involved in designing human experiments, collecting and analyzing data, running computational models, and comparing human and model performance.
Priority will be given to students who can commit to the project for two quarters. In addition to completing a final paper each quarter, students will be expected to present their research project at a lab meeting at the end of the second quarter.
How This Relates to Cognitive Science
The central goal of our research is to understand how humans learn and reason, and how intelligent machines might emulate these abilities. In both perception and reasoning, humans can often make surprisingly successful inferences from limited or imperfect data. The information available to us is frequently sparse (with very few examples), ambiguous (with multiple possible interpretations), and noisy (with a low signal-to-noise ratio). This ability to learn from sparse data stands in sharp contrast to current large-scale AI models, which typically rely on massive amounts of training data.
Humans and AI models sometimes achieve similar performance, but they can also differ substantially across tasks. Why do these similarities and differences arise? Do humans and AI models rely on similar or different representations? What factors shape the formation of these representations? Answering these questions can provide new insights into human cognition while also advancing the development of more intelligent and human-like AI systems. By combining cognitive science with computational modeling and modern AI techniques, our research seeks to understand the principles underlying learning, representation, and reasoning in both humans and machines.
Eye Tracking for Studying Memory and Cognition
- Faculty Sponsor: Maureen Ritchey
- Department: Psychology
- Contact Name: Maureen Ritchey
- E-mail: mritchey@ucla.edu
- Room Number: Psychology A349
- Website: www.thememolab.org
- Title of Rsearch: Eye Tracking for Studying Memory and Cognition
Description of Research Project
Eye movements can reveal how people attend to visual information, giving insight into what they will remember. We are running a series of eye-tracking studies examining how language and attention interact to influence memory. If interested, please submit an application here: https://forms.gle/qKAZ9z4ZfZigF7dRA
Description of Student Responsibilities
Students will assist with experimental design and data collection on a series of studies using eye-tracking methods to study memory and other cognitive processes. Other responsibilities may include scheduling participants, scoring participant responses, or developing experimental stimuli. Prior programming and/or data analysis experience is preferred. This is an in-person position within the Psychology building.
How This Relates to Cognitive Science
Through their involvement in a cognitive science research lab, students will develop skills in experimental design, human subjects data collection, and behavioral data analysis. Students will learn to operate the eye-tracker to obtain high-quality data. They will also read scientific research articles related to human memory and cognition.
Developmental Cognitive Signatures of Risk and Resilience in Psychosis
- Faculty Sponsor: Brittany Wolff
- Department: Psychiatry and Biobehavioral Sciences
- Contact Name: Brittany Wolff
- E-mail: bwolff@mednet.ucla.edu
- Room Number: C9-456
- Phone: (310) 447-2004
- Website: https://www.semel.ucla.edu/initiatives/seedlab/
- Title of Rsearch: Developmental Cognitive Signatures of Risk and Resilience in Psychosis
Description of Research Project
The SEED Laboratory (Signatures of Early Emergence and Divergence) investigates why individuals at familial and clinical high risk for psychosis and bipolar spectrum disorders follow different developmental trajectories, with some progressing to serious mental illness while others remain resilient. Supported by an NIMH K99/R00 Pathway to Independence Award, this research integrates neurocognitive, clinical, psychosocial, and multimodal longitudinal data to identify early developmental signatures that predict illness onset, remission, and resilience. Advanced statistical and machine learning approaches are used to characterize individual differences and translate these findings toward earlier identification and personalized intervention.
Description of Student Responsibilities
The student will contribute to ongoing SEED Laboratory and K99/R00 research through literature review, data organization and quality assurance, coding and management of neurocognitive and clinical data, and preparation of research materials. Depending on experience and project needs, the student may assist with neuropsychological data analysis, longitudinal datasets, data visualization, and computational or machine learning approaches. Responsibilities may also include participant recruitment for digital cognitive studies. The student will participate in laboratory meetings and scientific discussions and may contribute to conference presentations, manuscripts, and other research products.
How This Relates to Cognitive Science
This research directly examines cognition as a core component of neuropsychiatric development, including individual differences in attention, memory, executive functioning, processing speed, and variability in cognitive performance. It integrates cognitive psychology and neuropsychology with developmental science, neuroscience, computational modeling, and artificial intelligence to understand how cognitive processes relate to emerging psychopathology and resilience. The project therefore provides students with experience applying cognitive science methods and theory to clinically meaningful questions about human behavior, brain function, and mental health.