COGS202: Computational Modeling of Cognition

Course Description

This course surveys key methods and practices of computational modeling across a range of cognitive science fields. Bringing together basic tools of machine learning, AI, computational neuroscience, and probabilistic modeling, students will learn to use quantitative modeling to fit data, formulate hypotheses, and explain cognitive phenomena.

Learning Objectives

Read, comprehend, and critically evaluate primary research literature in computational cognitive science.
Identify different levels and styles of computational models.
Compare and contrast alternative computational modeling approaches.
Assess the strengths and potential pitfalls of computational models.
Design and implement computational modeling projects relevant to your own research questions within the cognitive sciences.

Materials and Course Structure

Each week will focus on a selected topic in computational modeling of cognition, representing a (biased) cross-section of key topics and modeling styles. Readings will include both papers for student presentations and additional readings.

Homework problems will provide practice with the basic concepts. Although students are encouraged to work in groups on homework, each student should write up their own solutions.

Presentations: Each student will present a research article once during the quarter (20 minutes presentation followed by 5 minutes of Q&A). Sign up for your presentation from the list in the tentative lecture schedule below at the following sign-up link.

Final Project: There will be a final project as part of the course requirements.

Grading Breakdown:

Attendance and participation: 5%
Homework (2 problem sets, first for 20% and second for 25%): 45% total
Paper presentation: 20%
Final project: 30%

Final Project Guidelines

COGS 202 Spring 2025

Presentation: In class on Week 10, Monday, June 2
Team Formation: Work in groups of 2-3 students. A Google Sheet for discussing ideas and forming groups is available here: Google Sheet
Proposal: A 1-page proposal is due May 13. The proposal should list the team members, the topic/tentative title, and 1-2 paragraphs about your initial thoughts on how you will approach the project.
Final Report: Due on June 6. The report should be roughly 5-8 pages (including references and figures). Your write-up should be organized as a scientific paper and include the following sections: Introduction (with a review of related work), Methods/Models, Results, and Discussion/Conclusion. For example, see:
Singh, P., Peterson, J. C., Battleday, R. M., & Griffiths, T. L. (2020). End-to-end deep prototype and exemplar models for predicting human behavior. Proceedings of the 42nd Annual Conference of the Cognitive Science Society [Link]. You can use the following template to structure your report (it is not required to stick to this template though )
Computational Work: The project must include some computational work (e.g., simulations, calculations, or computational analysis of experimental data).
Deliverable: A key deliverable is a set of specific, concrete predictions for the results of a proposed experimental test of the theory.
Project Format Options:
- Develop a computational model or analysis using your own research data.
- Reproduce and extend a result from one of the research papers we read in class.
- Develop a grant proposal for an experimental or computational project, including preliminary simulation or computational analysis validating the proposed idea.

Week	Date	Topic	Guest Lecture	Readings for Student Presentations	Additional Readings
Week 1	March 31	Introduction to Computational Models of Cognition	-	-	Marr Vision Chap 1 Embodied Turing Test What children can do that LLMs and LVMs cannot
Week 2	April 7	Statistical Framework for Modeling Brain/Behavioral Data (GLMs) / Coding Notebooks	-	Ten simple rules for the computational modeling of behavioral data How persuasive is a good fit? A comment on theory testing When a good fit can be bad	Encoding and decoding in fMRI
Week 3	April 14	GLM (ctd.) and Neural Networks/Deep Learning	Guest Lecture by Greta Tuckute: The Human Language Network and LLMs	Evaluating (and Improving) the Correspondence Between Deep Neural Networks and Human Representations Behavioral signatures of face perception in DNNs	Using ANNs to answer why questions about minds and brains NeuroAI Turing Test Towards an integration of deep learning and neuroscience
Week 4	April 21	Neural Networks/Deep Learning	Guest lecture by Jenelle Feather: Model metamers reveal divergent invariances between biological and artificial neural networks.	DNN models of sound localization reveal how perception is adapted to real-world environments Adapting deep neural networks to capture psychological representations	Qualitative similarities and differences in visual object representations between brains and DNNs
Week 5	April 28	Convolutional Neural Networks	-	Dissociating language and thought Language in brains, minds and machines Grounded language acquisition through the eyes and ears of a single child	LLMs as models of human cognition Meaning without reference in LLMs
Week 6	May 5	LLMs and Other Cognition	-	Do LLMs know what humans know? AI and the everything in the whole wide world benchmark Large Concept Models Social biases in word embeddings and their relation to human cognition.	-
Week 7	May 12	Probabilistic Models	-	Efficient inverse graphics in biological face processing How to Grow a Mind: Statistics, Structure, and Abstraction Probabilistic models of cognition	-
Week 8	May 19	Reinforcement Learning	-	RL in the brain Matching behavior and the representation of value in the parietal cortex What do reinforcement learning models measure? Interpreting model parameters in cognition and neuroscience	-
Week 9	May 26	Memorial Day Holiday	-	-	-
Week 10	June 2	Final Project Presentations	-	-	-

Course Policies

Attendance

This course is discussion-based, so attendance is essential for both benefiting from and contributing to class. While attendance is not directly graded, participation is. We understand that conferences and other events may lead to absences; please communicate any planned absences in advance. A limited number of absences will be allowed.

Collaboration

Discussion of class material is heavily encouraged. Collaboration on homework assignments is allowed as long as it is properly reported. Project collaboration is expected (projects are done in groups) and is also allowed across groups.

Academic Integrity

There is a zero tolerance policy for violations of academic integrity and course policies. If you have any doubts regarding the policies, please clarify with the course staff before proceeding.

Late Homework Policy

The maximum earnable points for each assignment will drop by 20% for each day an assignment is late.