Calendar

What is NLP for, and what does success look like?

W April 2
Introduction (slides)
Assignment 0 posted; read, sign, and submit (through Canvas) the academic integrity document
Case study on question answering (slides)
J&M 14 intro, 14.1, 14.4

Input-focused NLP -- classifying text

W April 9
Basics of classification, multinomial logistic regression, evaluation (slides)
Read J&M 5; Assignment 1 posted, due Monday April 21 Friday April 25
Neural networks (slides)
Read J&M 7 through 7.5
W April 16
Classification tasks and data (slides)
Assignment 2 posted, due Monday April 28 Wednesday April 30; read J&M 4 intro, 4.7, 4.8
Introductory generative models (slides)
Read J&M 4.1-5, 17.4

Segmenting text

W April 23
Data-driven segmentation of text; guest lecture from Alisa Liu (slides)
Read J&M 2 through 2.6; project checkpoint 1
Linguistic segmentation of text; guest lecture from Valentin Hofmann (slides)
Read Bender ch. 1 and 2;

Input-focused NLP -- embeddings

W April 30
Lexical semantics and embeddings (slides)
Read J&M 6 and optionally J&M 11 through 11.3; Assignment 3 posted, due Monday May 12

Output-focused NLP -- generating text

W April 30
Language models (slides)
Read J&M 3
W May 7
Language models, continued (slides)
Read J&M 7.6-7.7 and J&M chapter 8 through 8.7; Assignment 4 posted, due Monday, May 19; Project checkpoint 2
Machine translation, the canonical sequence-to-sequence task (slides)
Read J&M 13, skipping 13.4
W May 14
Attention and transformers (slides)
Read J&M 9; Assignment 5 posted, due May 26
Decoding (slides)
Read J&M 13.4
W May 21
Scaling up; guest lecture from Weijia Shi (slides)
Read J&M 10; Assignment 6 posted, due June 2; Project checkpoint 3
Prompting (slides)
Read J&M 12.1
W May 28
Finetuning, instructions, preferences (slides)
Read J&M 12, and J&M 14.3, and optionally J&M 11.4

Looking ahead to the future

W June 4
Societal impact of NLP; guest lecture from Hila Gonen (slides)
Open research challenges
W June 11
(no lecture)
Project due