CSCI 544 Fall 2026: Applied Natural Language Processing

Fall 2026 · Tuesday / Thursday, 4:00-5:50 PM · SAL 101 · 4 units

Course Staff

Teaching Assistants

TA office hours: Friday 11:00 AM-12:00 PM, GCS SB5 (under Level 2). Graders: TBD.

Course Information

Course-related questions: Use Piazza or ask in class or office hours. Email the instructor only for urgent personal matters; responses may take up to 48 hours. The instructor cannot provide D-clearance.

Download the Fall 2026 syllabus · Brightspace

Course Description

This course covers fundamental and cutting-edge topics in natural language processing, with a focus on language models. Students will study language modeling, deep learning, and machine learning; examine the capabilities and real-world consequences of large language models; and gain hands-on experience building and evaluating small-scale language models.

By the end of the course, students will be able to:

1. Apply modern NLP pipelines, including recurrent and Transformer-based sequence-to-sequence models.
2. Critically analyze contemporary research on large language models, including architectures, training and data paradigms, inference-time methods, and evaluation.
3. Explain the principles, assumptions, and trade-offs underlying modern LLM-based NLP systems.
4. Identify open research problems and lead scholarly discussions of recent top-tier research.
5. Design, execute, and clearly communicate an original graduate-level research project in applied NLP or large language models.

Course Schedule

Quiz and homework dates are subject to change.

Assignments and Grading

Homework Assignments (20% of total grade)

Through coding homework assignments, students will apply key pieces of modern natural language processing pipelines, such as recurrent and Transformer-based sequence-to-sequence models (Learning Objective O1). Homework coding assignments that will involve implementing core concepts using frameworks like PyTorch, then training and executing corresponding machine learning models on real natural language processing data. These coding assignments will be graded based on code correctness in terms of producing expected output, as well as through a written report documenting the code design choices and results of the training and evaluation experiments associated with the assignment.

There will be two coding homework assignments. The assignments must be done individually. Each assignment is graded on a scale of 0-100 and the specific rubric for each assignment is given in the assignment. Grading inquiries and questions about the grading of the homeworks and the quizzes can be asked (to the TAs) within two weeks from the grading date.

Paper review (10% of total grade):

Students will write a research paper review to explain concepts underlying natural language processing in their own words (Learning Objective O2). The course explores topics through a series of assigned readings in the form of book chapters. Also, the course project would require a literature review. Students will select one reading option and submit a two-page summary of all the reading. Reviews will be assessed based on answering a small set of questions, to be released at the time of the paper assignment, clearly and correctly. In most cases, each question will warrant at minimum a paragraph to answer.

Participation (10% of total grade)

Students are required to attend classes. Attendance will be taken at random on some days. Non-attendance can be the basis for lowering the grade.

Quizzes:

There will be six quizzes based on prior lectures. Missed quizzes will receive a zero grade, and there will be no make-up quizzes. Submit requests (form on Brightspace) for excused absences 24h prior to class start.

Grading Scale

Course final grades will be determined using the following scale:

Recommended Preparation

There are no formal prerequisites, but students should understand basic linear algebra, probability, and Python programming in a Linux environment. Fluency with Python is recommended.