# Schedule – 6.8610 Quantitative Methods for NLP – Advanced NLP

https://mit-6861.github.io/schedule

[![](https://mit-6861.github.io/images/mit_logo_resized.png)](https://mit-6861.github.io/)
# [6.8610 Quantitative Methods for NLP](https://mit-6861.github.io/)
Advanced NLP
[Home](https://mit-6861.github.io/) [Schedule](https://mit-6861.github.io/schedule)
* * *
**Spring 2026. Jacob Andreas, Chris Tanner, Omar Khattab**   
**T/TH 2:30pm-4:00pm in 54-100**
* * *
# Schedule  
| Date  | Topic  | Notes  |  
| --- | --- | --- |  
| **Tue, Feb 3**  | Intro + Classical NLP refresher  |   |  
| **Thu, Feb 5**  | Classical NLP / Word Representations  |   |  
| **Tue, Feb 10**  | seq2seq and Attention  |   |  
| **Thu, Feb 12**  | Transformers (MLM + decoder only)  | **HW1 released**  |  
| **Tue, Feb 17**  | [No class]  |   |  
| **Thu, Feb 19**  | Evaluation and desiderata for modern LMs  |   |  
| **Tue, Feb 24**  | Pretraining and Data Mixtures  |   |  
| **Thu, Feb 26**  | Scaling Laws and Emergent Behavior  | **HW1 due, HW2 released**  |  
| **Tue, Mar 3**  | Reasoning models & post-training paradigms  | **Pre-proposal due**  |  
| **Thu, Mar 5**  | Non-PG alignment algorithms: SLiC, DPO, KTO  |   |  
| **Tue, Mar 10**  | **Quiz 1**  |   |  
| **Thu, Mar 12**  | Policy-gradient RL: REINFORCE, GRPO, PPO  |   |  
| **Tue, Mar 17**  | Neural information retrieval  |   |  
| **Thu, Mar 19**  | Retrieval-augmented generation  | **HW2 due, HW3 released**  |  
| **Tue, Mar 24**  | [Spring break]  |   |  
| **Thu, Mar 26**  | [Spring break]  |   |  
| **Tue, Mar 31**  | Agents & inference scaling  | **April 1st: Project proposal due**  |  
| **Thu, Apr 2**  | LM programming & prompt optimization  |   |  
| **Tue, Apr 7**  | More than you ever wanted to know about tokenization  |   |  
| **Thu, Apr 9**  | Advanced architectures (MoE, Linear RNNs)  | **HW3 due**  |  
| **Tue, Apr 14**  | Efficient training and deployment  |   |  
| **Thu, Apr 16**  | Text diffusion  |   |  
| **Tue, Apr 21**  | **Quiz 2**  |   |  
| **Thu, Apr 23**  | Science of LMs (training dynamics)  |   |  
| **Tue, Apr 28**  | Safety  |   |  
| **Thu, Apr 30**  | Interpretability  |   |  
| **Tue, May 5**  | Human-AI interaction  |   |  
| **Thu, May 7**  | **Poster session**  |   |  
| **Tue, May 12**  | **Poster session**  | **Final Report due**  |