Course Materials & Downloads

History of NLP, Modern LLM Foundations & Paradigm Shifts

Word Representations, Distributed Semantics & Vector Embeddings (Word2Vec, GloVe)

Neural Network Fundamentals & Backpropagation in NLP

Autoregressive Language Modeling & Cross-Entropy Objectives

Recurrent Sequence Architectures: RNNs, LSTMs & Gated Recurrent Units (GRUs)

Attention Mechanisms & Transformer Architecture (RoPE, Multi-Head Attention)

Attention Mechanisms & Transformer Architecture (Supplementary / Extra Slides)

Quiz 1 Exam Paper (Covers Sessions S1–S6: Word Representations, Backprop, Language Modeling, RNNs & Transformers)

Subword Tokenization (BPE, SentencePiece) & Multilingual Language Models

LLM Pretraining: Web Data Pipelines, Quality Filtering & Scaling

Scaling Laws (Chinchilla) & In-Context Learning (ICL)

Parameter-Efficient Fine-Tuning (PEFT: LoRA, QLoRA) & Knowledge Distillation

Decoding Algorithms, Sampling Strategies & Chain-of-Thought (CoT) Prompting

Advanced Reasoning: System 2 Reasoning & Search

Retrieval-Augmented Generation (RAG) & Dense Passage Retrieval (DPR, Self-RAG)

Vision-Language Models (CLIP, LLaVA) & Multimodal Alignment

LLM Evaluation Methodologies, Benchmarks & Leaderboards (MMLU, LLM-as-a-Judge)

Graph Neural Networks for Natural Language Processing (GNN for NLP)

Diffusion Models & Flow Matching for Natural Language Generation

Reinforcement Learning Fundamentals in Natural Language Processing

Post-Training with Human Feedback & Preference Alignment (RLHF & DPO)

Language Model-Based Autonomous Agents, Function Calling & Tool Use (ReAct)

Model Quantization (AWQ, GPTQ, FP8/INT4) & Efficient Inference

Mixture of Experts (MoE) & Efficient Attention (FlashAttention-2)

Process Supervision & Test-Time Compute Scaling

Mechanistic Interpretability, Safety, Bias & Adversarial Attacks in LLMs

Research Skills, ACL Experimental Design & Open Research Frontiers in NLP

Language Models are Few-Shot Learners (GPT-3)

Tom B. Brown et al. (NeurIPS 2020)

Sparks of Artificial General Intelligence: Early experiments with GPT-4

Sébastien Bubeck et al. (Microsoft Research 2023)

Natural Language Understanding (Eisenstein Book, Chapter 1-2)

Jacob Eisenstein (MIT Press)

On the Opportunities and Risks of Foundation Models

Rishi Bommasani et al. (Stanford CRFM 2021)

Efficient Estimation of Word Representations in Vector Space (Word2Vec)

Tomas Mikolov, Kai Chen, Greg Corrado, Jeffrey Dean (ICLR 2013)

GloVe: Global Vectors for Word Representation

Jeffrey Pennington, Richard Socher, Christopher D. Manning (EMNLP 2014)

Distributed Representations of Words and Phrases and their Compositionality

Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg Corrado, Jeffrey Dean (NeurIPS 2013)

Enriching Word Vectors with Subword Information (fastText)

Piotr Bojanowski, Edouard Grave, Armand Joulin, Tomas Mikolov (TACL 2017)

Deep Learning (Review Paper)

Yann LeCun, Yoshua Bengio, Geoffrey Hinton (Nature 2015)

Yes you should understand backprop

Andrej Karpathy (2016)

Automatic Differentiation in PyTorch

Adam Paszke et al. (NeurIPS Autodiff Workshop 2017)

The spelled-out intro to neural networks and backpropagation: building micrograd

Andrej Karpathy (YouTube Lecture)

Neural Networks Pt. 2: Backpropagation Main Ideas

StatQuest with Josh Starmer

Backpropagation, intuitively | Deep Learning Chapter 3

3Blue1Brown (Grant Sanderson)

A Neural Probabilistic Language Model

Yoshua Bengio, Réjean Ducharme, Pascal Vincent, Christian Jauvin (JMLR 2003)

Generating Sequences With Recurrent Neural Networks

Alex Graves (2013)

Speech and Language Processing (3rd ed. draft, Chapter 3: N-gram Language Models)

Dan Jurafsky, James H. Martin

Long Short-Term Memory

Sepp Hochreiter, Jürgen Schmidhuber (Neural Computation 1997)

Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling

Junyoung Chung, Caglar Gulcehre, Kyunghyun Cho, Yoshua Bengio (NeurIPS 2014)

Understanding LSTM Networks

Christopher Olah (2015)

The Unreasonable Effectiveness of Recurrent Neural Networks

Andrej Karpathy (2015)

Recurrent Neural Networks (RNNs), Clearly Explained!!!

StatQuest with Josh Starmer

Long Short-Term Memory (LSTM), Clearly Explained

StatQuest with Josh Starmer

Attention Is All You Need

Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit et al. (NeurIPS 2017)

The Illustrated Transformer

Jay Alammar (2018)

RoFormer: Enhanced Transformer with Rotary Position Embedding (RoPE)

Jianlin Su, Yu Lu, Shengfeng Pan, Ahmed Murtadha, Bo Wen, Yunfeng Liu (Neurocomputing 2024)

The Annotated Transformer

Austin Rush et al. (Harvard NLP)

Neural Machine Translation of Rare Words with Subword Units (BPE)

Rico Sennrich, Barry Haddow, Alexandra Birch (ACL 2016)

Subword Regularization: Improving Neural Network Translation Models (SentencePiece / Unigram)

Taku Kudo (ACL 2018)

Cross-lingual Language Model Pretraining (XLM)

Guillaume Lample, Alexis Conneau (NeurIPS 2019)

Language Models are Unsupervised Multitask Learners (GPT-2)

Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever (OpenAI 2019)

The Llama 3 Herd of Models

Abhimanyu Dubey et al. (Meta AI 2024)

The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora

Guilherme Penedo et al. (NeurIPS 2023)

Training Compute-Optimal Large Language Models (Chinchilla)

Jordan Hoffmann et al. (DeepMind, NeurIPS 2022)

Scaling Laws for Neural Language Models

Jared Kaplan et al. (OpenAI 2020)

LoRA: Low-Rank Adaptation of Large Language Models

Edward J. Hu et al. (ICLR 2022)

QLoRA: Efficient Finetuning of Quantized LLMs

Tim Dettmers, Artidoro Pagnoni, Ari Holtzman, Luke Zettlemoyer (NeurIPS 2023)

Distilling the Knowledge in a Neural Network

Geoffrey Hinton, Oriol Vinyals, Jeff Dean (NeurIPS Deep Learning Workshop 2014)

Prefix-Tuning: Optimizing Continuous Prompts for Generation

Xiang Lisa Li, Percy Liang (ACL 2021)

Chain-of-Thought Prompting Elicits Reasoning in Large Language Models

Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc Le, Denny Zhou (NeurIPS 2022)

The Curious Case of Neural Text Degeneration (Nucleus / Top-p Sampling)

Ari Holtzman, Jan Buys, Li Du, Maxwell Forbes, Yejin Choi (ICLR 2020)

Tree of Thoughts: Deliberate Problem Solving with Large Language Models

Shunyu Yao et al. (Princeton & DeepMind, NeurIPS 2023)

STaR: Bootstrapping Reasoning With Reasoning

Eric Zelikman et al. (Stanford, NeurIPS 2022)

Large Language Monkeys: Scaling Inference Compute with Repeated Sampling

Bradley Brown et al. (2024)

Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks (RAG)

Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni et al. (NeurIPS 2020)

Dense Passage Retrieval for Open-Domain Question Answering (DPR)

Vladimir Karpukhin, Barlas Oğuz, Sewon Min, Patrick Lewis, Ledell Wu, Sergey Edunov, Danqi Chen, Wen-tau Yih (EMNLP 2020)

Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection

Akari Asai, Zeqiu Wu, Yizhong Wang, Avirup Sil, Hannaneh Hajishirzi (ICLR 2024)

Learning Transferable Visual Models From Natural Language Supervision (CLIP)

Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal et al. (OpenAI, ICML 2021)

Visual Instruction Tuning (LLaVA)

Haotian Liu, Chunyuan Li, Qingyang Wu, Yong Jae Lee (NeurIPS 2023)

An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale (ViT)

Alexey Dosovitskiy et al. (ICLR 2021)

Measuring Massive Multitask Language Understanding (MMLU)

Dan Hendrycks et al. (ICLR 2021)

Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena

Lianmin Zheng, Wei-Lin Chiang, Hao Zhang, Siyuan Zhuang, Yonghao Zhuang, Lingxiao Zheng et al. (NeurIPS 2023)

Adding Error Bars to Evals: A Statistical Approach to Language Model Evaluations

Noah A. Smith et al. (2024)

Graph Neural Networks for Natural Language Processing: A Survey

Lingfei Wu, Yu Chen, Kai Shen, Xiaojie Guo et al. (Foundations and Trends in ML 2021)

Graph Convolutional Networks for Text Classification

Liang Yao, Chengsheng Mao, Yuan Luo (AAAI 2019)

Semi-Supervised Classification with Graph Convolutional Networks (GCN)

Thomas N. Kipf, Max Welling (ICLR 2017)

Diffusion-LM Improves Controllable Text Generation

Xiang Lisa Li, John Thickstun, Ishaan Gulrajani, Percy Liang, Tatsunori B. Hashimoto (NeurIPS 2022)

Flow Matching for Generative Modeling

Yaron Lipman, Ricky T. Q. Chen, Heli Ben-Hamu, Maximilian Nicklas, Matt Le (ICLR 2023)

Denoising Diffusion Probabilistic Models (DDPM)

Jonathan Ho, Ajay Jain, Pieter Abbeel (NeurIPS 2020)

Deep Reinforcement Learning from Human Preferences

Paul Christiano, Jan Leike, Tom B. Brown, Miljan Martic, Shane Legg, Dario Amodei (NeurIPS 2017)

Proximal Policy Optimization Algorithms (PPO)

John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, Oleg Klimov (OpenAI 2017)

Training language models to follow instructions with human feedback (InstructGPT)

Long Ouyang et al. (OpenAI, NeurIPS 2022)

Direct Preference Optimization: Your Language Model is Secretly a Reward Model (DPO)

Rafael Rafailov, Archit Sharma, Eric Mitchell, Stefano Ermon, Christopher D. Manning, Chelsea Finn (NeurIPS 2023)

ReAct: Synergizing Reasoning and Acting in Language Models

Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, Yuan Cao (ICLR 2023)

Toolformer: Language Models Can Teach Themselves to Use Tools

Timo Schick, Jane Dwivedi-Yu, Roberto Dessì, Roberta Raileanu, Maria Lomeli, Luke Zettlemoyer, Nicola Cancedda, Thomas Scialom (NeurIPS 2023)

World of Bits: An Open-Domain Platform for Web-Based Agents

Tianlin Shi, Andrej Karpathy, Linxi Fan, Jonathan O. Hernandez, Percy Liang (ICML 2017)

LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale

Tim Dettmers, Mike Lewis, Younes Belkada, Luke Zettlemoyer (NeurIPS 2022)

AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration

Ji Lin, Jiaming Tang, Haotian Tang, Shang Yang, Xingyu Dang, Song Han (MLSys 2024)

ZeRO: Memory Optimizations Toward Training Trillion Parameter Models

Samyam Rajbhandari, Jeff Rasley, Olatunji Ruwase, Yuxiong He (SC 2020)

Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer

Noam Shazeer, Azalia Mirhoseini, Krzysztof Maziarz, Andy Davis, Quoc Le, Geoffrey Hinton, Jeff Dean (ICLR 2017)

FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning

Tri Dao (ICLR 2024)

Mixtral of Experts

Albert Q. Jiang et al. (Mistral AI 2024)

Let's Verify Step by Step (PRM800K & Process Supervision)

Hunter Lightman et al. (OpenAI 2023)

Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters

Charlie Snell et al. (Google DeepMind / UC Berkeley 2024)

Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking

Eric Zelikman et al. (Stanford 2024)

Towards Monosemanticity: Decomposing Language Models With Dictionary Learning

Trenton Bricken et al. (Anthropic 2023)

Universal and Transferable Adversarial Attacks on Aligned Language Models

Andy Zou, Zifan Wang, Nicholas Carlini, Milad Nasr, J. Zico Kolter, Matt Fredrikson (2023)

Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)

Been Kim, Martin Wattenberg, Justin Gilmer, Carrie Cai, James Wexler, Fernanda Viegas, Rory Sayres (ICML 2018)

A Survey of Large Language Models

Wayne Xin Zhao et al. (AI Open 2023)

Holistic Evaluation of Language Models (HELM)

Percy Liang et al. (Stanford CRFM, TMLR 2023)

Introduction & Environment Setup in PyTorch

Hands-on setup with PyTorch, Hugging Face Transformers, and basic neural tensor operations.

Word Representations, Word2Vec & Text Classification

Implementing word vector embeddings, continuous representations, and sentiment classification.

Autoregressive Neural Language Modeling Fundamentals

Building next-token prediction language models from scratch with Cross-Entropy Loss.

Recurrent Neural Networks (RNN & LSTM) in PyTorch

Implementing recurrent sequence encoders, LSTM cells, and gated hidden state transitions.

Transformer Architecture & Scaled Dot-Product Attention

Modular implementation of Multi-Head Self-Attention, Positional Encoding, and Feed-Forward blocks.

LLM Pretraining Pipelines & Data Preprocessing at Scale

Web text scraping, quality filtering, deduplication pipelines, and batch sequence packing.

In-Context Learning (ICL) & Few-Shot Prompt Engineering

Demonstrating few-shot learning, prompt calibration, and instruction formatting across open LLMs.

Instruction Fine-Tuning & Parameter-Efficient LoRA Adaptation

Fine-tuning foundation models with LoRA/QLoRA using Hugging Face PEFT and TRL.

Decoding Algorithms, Sampling (Top-p/Top-k) & Search

Greedy decoding, temperature scaling, top-k/nucleus sampling, and beam search implementations.

Retrieval-Augmented Generation (RAG) & Vector Indexing

Dense passage retrieval (DPR), vector database indexing with FAISS, chunking, and generation.

Multimodal Vision-Language Models (CLIP & Visual Embeddings)

Contrastive image-text training (CLIP), zero-shot visual classification, and multimodal alignment.

Multimodal Instruction Tuning & Visual Generation (LLaVA)

Cross-modal projection, Visual Question Answering (VQA), and multimodal chat assistants.

LLM Evaluation Benchmarks & Automated Scoring Metrics

Benchmarking with MMLU and GSM8K, LLM-as-a-Judge evaluations, and statistical confidence intervals.

Research Skills, Experimental Design & Ablation Analysis

Designing rigorous NLP ablation experiments, statistical significance testing, and artifact analysis.

Diffusion Models & Flow Matching for Text Generation

Implementing continuous/discrete text diffusion models, denoising steps, and flow matching.

Reinforcement Learning Fundamentals & Policy Optimization (PPO)

Markov Decision Processes, policy gradients, advantage estimation, and Proximal Policy Optimization.

Preference Alignment for LLMs (RLHF & Direct Preference Optimization / DPO)

Bradley-Terry reward modeling, RLHF with PPO, and Direct Preference Optimization (DPO).

LLM-Based Autonomous Agents, Function Calling & Tool Use (ReAct)

Building autonomous agent loops with ReAct prompting, API tool integration, and environment feedback.

Model Quantization (AWQ, GPTQ & INT8/INT4 Precision)

Weight quantization with bitsandbytes, activation-aware quantization (AWQ), and inference benchmarking.

Distributed Training, Model Parallelism & ZeRO Optimization

Tensor and pipeline parallelism (Megatron-LM), ZeRO memory stages (DeepSpeed), and FSDP on GPU clusters.

MMLU (Massive Multitask Language Understanding)

57 subjects across STEM, humanities, and social sciences for holistic LLM benchmarking.

GSM8K (Grade School Math 8K)

8,500 high-quality multi-step math reasoning word problems for Chain-of-Thought evaluation.

UltraFeedback & Anthropic HH-RLHF

Binarized pairwise human preference datasets for RLHF and Direct Preference Optimization (DPO).

MS MARCO & HotpotQA

Passage ranking and multi-hop reasoning datasets for open-domain RAG benchmarking.

ANLP Course Pack v1 (Project Multi-Hop Dataset)

3,882 Wikipedia passages, 300 dev questions with evidence triples & labels, and 300 hidden test questions for course project.

ANLP Project Starter Kit (score.py, config & baseline)

Official starter kit archive containing score.py with bootstrap CI, example_predict.py, and predictions format specification.