Schedule

1. 01 · Sessions 1–5
   Perception How models understand what they see and read: the basics of networks and Transformers, then seeing images and linking them with text.
2. 02 · Sessions 6–7
   Generation How models create new content: generating images, then videos, from a text prompt.
3. 03 · Sessions 8–9
   Multimodal LLMs How one model brings the two together: scaled-up LLMs, then multimodal LLMs that unify perception and generation.

See also Stanford CS231n Deep Learning for Computer Vision (Spring 2025) and MIT 6.S978 Deep Generative Models (Fall 2024).

1. 01 · Sessions 1–5
   Training LLMs How an LLM is trained and used: pretraining, then prompting that puts it to good use, and post-training that aligns it with what people want.
2. 02 · Sessions 6–9
   From LLMs to Agents How an LLM grows into an agent: reasoning over many steps, memory beyond a single context, then acting in a loop with tools.

See also Princeton COS 597R Deep Dive into Large Language Models (Fall 2024).

1. 01 · Sessions 1–3
   Architectures and Training How deep networks are built and trained: backpropagation first, then architectures with the right inductive bias for each data type.
2. 02 · Sessions 4–7
   Representation and Generation How networks model data: they encode it into representations and decode them into new samples.
3. 03 · Sessions 8–10
   Generalization and Theory How models generalize beyond the training data: to shifted distributions and new tasks, and why this works at all.

Based on MIT 6.7960 Deep Learning (Fall 2025).

1. 01 · Sessions 1–6
   Robot Embodiment How robot tasks mirror human abilities: arms that grasp, legs that walk, a sense of where to go, and humanoid bodies with dexterous hands, all learned from data.
2. 02 · Sessions 7–11
   Foundation Models for Robotics How foundation models help robots act: an understanding of 3D space, vision and language turned into action, and world models that bring perception and planning.

See also Berkeley CS 294-277 Robots That Learn (Spring 2026) and Berkeley CS 294-43 Large Scale Vision, Language, and Action Models (Fall 2025).