Word-object learning via visual exploration in space (WOLVES)

Overview of WOLVES

This case study is provides an overview of WOLVES, a DF model of word learning reported in Bhat, Spencer & Samuelson (2022) Psychological Review. Here's a link to the paper:


Here's an introductory lecture on the area of cross-situational word learning and the WOLVES model.

If you'd like to run the model, you'll need to download COSIVINA. Instructions for that are below along with instructions for downloading the code from our github site.

Setting up COSIVINA

To play around with the model from the paper, you will first need to set up the COSIVINA simulation environment. You can clone the COSIVINA repository from https://github.com/cosivina. For an overview of how to do this, see my video tour which includes details of how to use the GitHub Desktop App along with our DF respositories. For details, see:

In addition to COSIVINA, you'll need the simulator from the ‘WOLVES_PsychReview_2021” folder at: https://github.com/cosivina/cosivina_dft_projects

You'll also need to download the jsonlab support files: https://github.com/fangq/jsonlab

Running the Simulator

To play around with the model, launch matlab. Next, navigate to your COSIVINA folder and type 'setpath'. This will link in the core COSIVINA files and jsonlab support. Next, go to the ‘WOLVES_PsychReview_2021' folder and add 'selected folders and subfolders' to the matlab path by right-clicking on the folder. Then, navigate to the 'wolves_core' folder and run the ‘XSIT_Manual_run.m' file in matlab. Comments in this file explain how to run the simulator.

Here's a quick video demo of these steps:


For a step-by-step tour and discussion of the model, see the video below recorded during the Dynamic Friday Tutorial on December 2, 2022.