UNIVERSITY BREAKS AND IMPORTANT DATES
Please visit the Harvard 10-year Calendar for the latest dates:
- Sept 1 (Wed): Semester begins
- Sept 6 (Mon): Labor Day (no classes)
- Oct 11 (Mon): Indigenous People’s Day (no classes)
- Oct 20 (Wed): Kamala Harris’ birthday (classes are held)
- Nov 11 (Thurs): Veteran’s Day (classes are held)
- Nov 24 (Wed) - Nov 28 (Sun): Thanksgiving Break
- Dec 2 (Thurs): Last day of classes/lectures
- Dec 3 (Wed) - Dec 8 (Wed): Reading Period (no classes)
- Dec 9 (Thurs) - Dec 18 (Sat): Exam Period
OUR CLASS
- Lectures are on Tuesdays and Thursdays at 9:45am-11:00am in SEC LL2.221
- first lecture is on Sept 2 (Thurs). Mandatory attendance for anyone wanting to enroll.
- earlier lectures will mostly concern Deep Learning models
- later lectures will mostly concern NLP Tasks
- the last content-based lecture will be Nov 11 (Thurs)
- the Exam will be given during the following lecture time, Nov 16 (Tues)
- four lectures after this are designated for Open Project Discussions, whereby we will check-in with Research Project groups and offer assistance and feedback
- we will not have class on Nov 25 (Thurs) and Dec 7 (Tues), due to university breaks
- Final Presentations (pre-recorded videos) will be on Dec 9 (Thurs)
- ~8 unannounced pop quizzes will be issued, each of which in-class at the beginning of a lecture.
- Homeworks will be released on Tuesdays and will be due two weeks later (Mondays at 11:59pm EST).
- Research Projects will span 12 weeks of the semester, with several deliverables due throughout:
- Phase 1 is due Sept 30 (Thurs) @ 11:59pm EST
- Phase 2 is due Oct 14 (Thurs) @ 11:59pm EST
- Phase 3 is due Oct 28 (Thurs) @ 11:59pm EST
- Phase 4 is due Nov 11 (Thurs) @ 11:59pm EST
- Phase 5 is due Nov 23 (Tues) @ 11:59pm EST
- Final Presentations (pre-recorded videos) will be on Dec 9 (Thurs)
LECTURES
- Lecture 1: Introduction + Logistics [pdf] [pptx]
- Lecture 2: Representations + BoW + TFIDF [pdf] [pptx]
- Lecture 3: Language Modelling + ngrams [pdf] [pptx]
- Lecture 4: Neural Language Models [pdf] [pptx]
- Lecture 5: Recurrent Neural Networks [pdf] [pptx]
- Lecture 6: LSTMs [pdf] [pptx]
- Lecture 7: seq2seq [pdf] [pptx]
- Lecture 8: Machine Translation [pdf] [pptx]
- Lecture 9: Self-Attention [pdf] [pptx]
- Lecture 10: Transformers [pdf] [pptx]
- Lecture 11: BERT [pdf] [pptx]
- Lecture 12: GPT-2 [pdf] [pptx]
- Lecture 13: Embedded Bias (guest lecture) [pdf] [pptx]
- Lecture 14: Summarization [pdf] [pptx]
- Lecture 15: Entity Linking [pdf] [pptx]
- Lecture 16: Coreference Resolution [pdf] [pptx]
- Lecture 17: Commonsense [pdf]
- Lecture 18: Adversarial NLP [pdf] [pptx]
- Lecture 19: Interpretability and Probing [pdf] [pptx]
- Lecture 20: Review [pdf] [pptx]
- Lecture 21: Exam Study Session [pdf]
SPREADSHEET FORMAT
- The
HW #
column denotes which lectures are eligible for comprising the content for each homework (i.e., each homework encompasses four lectures) - Lecture topics with a white background represent foundational content and Deep Learning models
- Lecture topics with a gray background represent NLP tasks
- Lecture topics with an orange background represent guest lectures