Welcome, Benchmarkers!
Benchmarks are at the core of Brain-Score. They are used to test how well models match experimental observations. If you want to add your experimental results into Brain-Score as a benchmark, the process involves two main steps: first, you need to add your experimental stimuli as well as subject data. This is covered in the first section of this tutorial, "Package Data".
The second step involves creating and defining the benchmark file itself: This is how and where you choose benchmark hyperparameters, such as visual degrees, a metric, and a ceiling. This is covered in the "Create a Benchmark" section of this tutorial.
Note: Brain-Score supports three types of benchmarks, differentiated by the type of experimental data that was gathered: neural (fMRI, PSTH, EEG, etc), behavioral, and engineering. In this benchmark tutorial series, we will explore a sample behavioral benchmark, Ferguson2024.
Our tutorials and FAQs, created with Brain-Score users, aim to cover all bases. However, if issues arise, reach out to our community or consult the troubleshooting guide below for common errors and solutions.