The Minimal Structure of Deep Learning | MIT 6.S191 Lecture 1
Lecture 1 of the 2026 course builds the vocabulary shared by the rest of the course: perceptrons, forward propagation, loss, and gradient descent.
#6s191
12 postsSequence Modeling: From RNNs to Attention | MIT 6.S191 Lecture 2
Lecture 2 of the 2026 course addresses data where order changes meaning—text, audio, and time series—and connects directly to music generation in Lab 1.
Computer Vision: How Convolution Preserves Spatial Structure | MIT 6.S191 Lecture 3
Lecture 3 of the 2026 course moves from image tensors, convolution, and pooling to recognition systems, preparing for MNIST and face detection in Lab 2.
Generative Modeling: From Latent Spaces to Diffusion | MIT 6.S191 Lecture 4
Lecture 4 of the 2026 course separates generative from discriminative tasks, organizes VAE, GAN, and diffusion objectives, and leads into Lab 2’s DB-VAE.
Reinforcement Learning: Learning from Return Instead of Labels | MIT 6.S191 Lecture 5
Lecture 5 of the 2026 course connects agent, environment, state, action, reward, and policy into an interaction loop, introducing credit assignment and exploration.
New Frontiers: Choosing the Problem Beyond the Model | MIT 6.S191 Lecture 6
Lecture 6 of the 2026 course places deep learning in emerging applications and real constraints, emphasizing data, outputs, evaluation, and failure conditions.
The Three Laws of AI: Safety Through Observability and Evaluation | MIT 6.S191 Lecture 7
Lecture 7 of the 2026 course starts from Asimov’s literary laws and examines modern safety protocols through traces, test data, and continuous evaluation.
AI for Science: Putting Domain Structure into Learning | MIT 6.S191 Lecture 8
Lecture 8 of the 2026 course uses the scientific-discovery loop to show how simulators, AI emulators, and experiments cooperate instead of reducing science to generic prediction.
Massively Parallel Training: Memory and Communication Set the Boundary | MIT 6.S191 Lecture 9
Lecture 9 of the 2026 course starts with GPU memory pressure and moves through checkpointing, offloading, ZeRO, FSDP, and multiple forms of parallelism.
Generate Music with PyTorch and an LSTM | MIT 6.S191 Lab 1
In the 2026 lab, students cover tensors, autograd, and modules before turning ABC notation into character sequences for LSTM music generation.
From MNIST to Facial Debiasing with a DB-VAE | MIT 6.S191 Lab 2
In the 2026 lab, part 1 classifies MNIST with dense and convolutional networks; Part 2 learns a facial latent distribution with a DB-VAE and changes training sampling.
LoRA Fine-Tuning and LLM-as-a-Judge Evaluation | MIT 6.S191 Lab 3
In the 2026 lab, students build chat templates and generation with LFM2-1.2B, adapt style through LoRA, and combine OpenRouter with Opik for a judge workflow.