Archive

These earlier resources are retained because they may still be useful and because existing publications and links refer to them. Unless noted otherwise, they are no longer actively maintained.

Literature and guides

Papers comparing brains, behavior, and deep neural networks

A collaboratively edited literature list assembled during the rapid growth of work comparing deep neural networks with human behavior and brain activity. The list was maintained through approximately 2022–2023 and is no longer comprehensive.

Guide to running Amazon Mechanical Turk studies

Practical notes on setting up online experiments, improving data quality, and managing costs on Mechanical Turk. Platform details may have changed since the guide was written.

Stimulus-generation examples

Continuous flash suppression exampleContinuous flash suppression masks

Continuous flash suppression presents a target image to one eye and rapidly changing Mondrian masks to the other. The original Matlab code for generating the masks remains available for download.

Phase-scrambled mandrill imagePhase-scrambled images

Phase scrambling can preserve low-level image properties while reducing the visibility of image content, for example when constructing baseline stimuli for object-localizer experiments.

Kermit embedded in an image cloudImage clouds and luminance illusions

This function generates image clouds that can be used to construct displays such as the Anderson and Winawer luminance illusion.

Earlier Matlab functions

MRI movie exampleNIfTI movie and motion inspection

A tool for rapidly displaying movies of raw images, difference images, and activation images across time. The code is old but remains available for compatibility with existing workflows.

Experimental design efficiencyfMRI design-efficiency tools

Functions for inspecting and optimizing the efficiency of event-related fMRI designs across candidate stimulus orders and timing parameters.

Random-walk value chainsUncorrelated random walks

A Matlab function for generating random-walk value chains with comparable spectral properties for learning and decision-making experiments.

Disclaimer