Announcements

Teaching Team

Grading Policy

Note: more details regarding the grading policy will be discussed in the first lecture.

• Attendance & Participation: 8% (+2% survey)
• Assignments (Individual): 30% (10% each)
• In-class Quizzes: 30% (15% each)
• Final (Group) Project: 30%

Class Information

Lectures: Room 2.505

• Monday: 1:00PM – 3:00PM
• Tuesday: 1:30PM – 3:30PM
• Note: No recording or streaming, please come!

Cohorts: Room 1.416

• CI01: Tuesday 9:30AM – 10:30AM
• CI02: Thursday 4:00PM – 5:00PM
• Note: Cohorts only in certain weeks, check schedule below

Course Schedule

Note: Schedule is tentative and subject to change!
Materials: Lecture slides and cohort materials can be found on Edimension platform.

Week	Date	Topics	Materials / Readings	Cohorts	Deadlines
1	15-Sep	Course Logistics Introduction to NLP		-	-
2	22-Sep	Recap on ML/NN Word Vectors	1. Stanford CS231n note on neural network basics and backpropagation 2. Efficient Estimation of Word Representations in Vector Space (original word2vec paper) 3. The Illustrated Word2vec 4. word2vec Parameter Learning Explained	-	Release A1
3	29-Sep	Words and Tokens Language Model	1. BPE tutorial by Hugging Face 2. N-gram Language Models	Word2Vec Tutorial	A1 Due
4	6-Oct	Quiz-1 RNN and Variants	1. The Unreasonable Effectiveness of Recurrent Neural Networks 2. Understanding LSTM Networks	-	-
5	13-Oct	Seq2Seq & Attention 1 Seq2Seq & Attention 2	1. Sequence to Sequence (seq2seq) and Attention (with good visualizations)	-	Release A2
6	20-Oct	Practical Tips of NLP Project Project Announce	1. Practical Methodology (Deep Learning book chapter)	Transformer Tutorial	A2 Due
7	27-Oct	Recess Week
8	3-Nov	Transformer LLM Pre-training	1. Attention Is All You Need 2. The Illustrated Transformer 3. The Annotated Transformer 4. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding	-	-
9	10-Nov	Project Proposal Presentation Post-training (SFT/RLHF)	1. Aligning language models to follow instructions (InstructGPT) 2. Scaling Instruction-Finetuned Language Models (Flan-T5)	LLM Tutorial	-
10	17-Nov	Efficient Adaptation Quiz-2	1. Language Models are Few-Shot Learners 2. Chain-of-Thought Prompting Elicits Reasoning in Large Language Models 3. LoRA: Low-Rank Adaptation of Large Language Models	-	Release A3
11	24-Nov	Evaluation Guest Lecture on test-time compute and RL	-	Help on Project	A3 Due
12	1-Dec	Advanced Topic-3 Advanced Topic-4	-	-	-
13	8-Dec	Final Project Presentations