On August 24, Class 4 was held. In this class, we worked on how language models learn, the meanings they process, and where they acquire that knowledge. We also discussed the social risks that these models might bring. What can we say about bias and polarization?
We also carried out a series of activities. We brought some exercises that can be worked on in classrooms using unplugged templates, offline, but also how to later transfer them to EDIA.
Next week, we will be sending an asynchronous activity that is part of the practice and final project.
We will also open a space for virtual consultations before the next in-person meeting. While not mandatory, this will offer the possibility to access specific questions if needed.
We are nearing the final stretch! We hope you’re enjoying the process!
Teaching Material of Interest:
Read an article from El País: “Two Latin American Academics, Argentina’s Luciana Benotti and Chile’s Jocelyn Dunstan, Explain Why Spanish-Speaking Efforts Are Important to Expand Language Models Available in Languages Other Than English.” Link here.
