C S 574
Download as PDF
Transformer Models for Natural Language Processing
Computer ScienceCollege of Computational, Mathematical, & Physical Sciences
Course Description
Understand and learn how to use state of the art transformer language models for many Natural Language Processing and Understanding tasks, such as text classification, text generation, question answering, machine translation, and many other tasks.
When Taught
Fall
Fixed/Max
3
Fixed
3
Fixed
0
Title
Architectural Foundations
Learning Outcome
Construct and explain the internal mechanics of transformer architectures including self-attention mechanisms and positional encoding. This fulfills the Intellectually Enlarging aim by requiring the rigorous mental discipline and mathematical depth needed to master state-of-the-art models.
Title
Model Implementation
Learning Outcome
Develop and fine-tune large-scale language models for specific downstream tasks such as abstractive summarization or sentiment analysis. This outcome is Character Building because it demands "sustained effort" and professional integrity when handling massive datasets and intensive computational resources.
Title
Ethical Deployment
Learning Outcome
Evaluate the ethical implications of generative AI including algorithmic bias and truthfulness to propose responsible deployment strategies. This aligns with the Spiritually Strengthening aim by encouraging students to use technology in ways that protect human dignity and honor the agency of all individuals.
