-
- Homepage
- https://cme296.stanford.edu/
- About
- The Amidi brothers are known for creating the widely-used machine learning cheatsheets for Stanford's CS229 and CS230 courses, now used by millions of learners worldwide.
- Topics
- Diffusion models · Score matching and flow matching · Diffusion Transformers and U-Nets · Controllable image generation · Model evaluation · Video generation
- Notes
- An 8-lecture Stanford graduate course on diffusion-based generative models for vision, covering the full stack from DDPM and score matching through modern architectures like U-Nets and Diffusion Transformers. Includes controllable generation, model training and finetuning, and evaluation metrics. Lecture slides are released publicly alongside YouTube recordings as the Spring 2026 course progresses.
-
- Homepage
- https://president.yale.edu/committees-programs/devane-lectures/america-at-250-a-history
- Topics
- U.S. political history 1776–present · Race and Reconstruction · Cold War and national security · American identity
- Notes
- One-time-only Yale course for the nation’s 250th anniversary, asking what America is and was meant to be. Three eminent historians, each with a distinct lens. Weekly post-lecture discussions by the professors are also posted to YouTube.
-
- Homepage
- https://stanford-cs221.github.io/autumn2025/
- About
- Directs Stanford’s Center for Research on Foundation Models (CRFM); known for benchmarking LLMs.
- Topics
- Machine learning · Search · Markov decision processes · Bayesian networks · Logic · Language models
- Notes
- Stanford’s flagship AI course, rigorous and broad.
-
- Homepage
- https://cs336.stanford.edu/
- About
- Liang directs Stanford's Center for Research on Foundation Models (CRFM); Hashimoto works on language model evaluation and robustness.
- Topics
- Tokenization · Transformer architectures · GPU kernels · Parallelism · Scaling laws · LLM evaluation
- Notes
- A systems-level course that builds a language model from scratch — tokenizer, attention, GPU kernels, data pipelines, and scaling laws. Unusually implementation-heavy even for a Stanford graduate course, with all five assignments on GitHub. Full 2025 lecture series released publicly on YouTube.
-
- Homepage
- https://optimalcontrol.ri.cmu.edu/
- About
- Works on trajectory optimization, spacecraft dynamics, and fast numerical methods for robot motion planning.
- Topics
- LQR · Trajectory optimization · iLQR and DDP · State estimation · System identification · Reinforcement learning
- Notes
- A graduate robotics course on controlling real physical systems: covers classical optimal control (Pontryagin, Riccati, LQR), numerical trajectory optimization, and how these connect to modern RL. Strong on both theory and implementation; lecture notes and Jupyter notebooks are posted with each lecture.
-
- Homepage
- https://ocw.mit.edu/courses/18-156-projection-theory-spring-2025/
- About
- Known for major work in geometric combinatorics and harmonic analysis.
- Topics
- Projection theorems · Geometric measure theory · Additive combinatorics · Harmonic analysis · Homogeneous dynamics
- Notes
- A recent graduate analysis course built around a field that had major breakthroughs very recently. The OCW release is unusually complete: full lecture videos, polished notes, and weekly problem sets.
-
- Homepage
- https://ocw.mit.edu/courses/18-100b-real-analysis-spring-2025/
- About
- Geometric analyst known for work on minimal surfaces and geometric PDE.
- Topics
- Real numbers · Proof techniques · Continuity · Differentiation · Riemann integration
- Notes
- Fresh 2025 OCW capture of MIT's core analysis sequence, with the standard epsilon-delta backbone presented in full lecture-video form. A strong seed entry for anyone wanting a rigorous modern baseline in pure math.
-
- Homepage
- https://cs50.harvard.edu/extension/sql/2025/spring/
- Topics
- Relational databases · SQL querying · Schema design · Views and CTEs · Indexes · Scaling
- Notes
- A clean, focused databases course with on-demand lecture videos and a full progression from basic querying through indexing and replication. More practical than theoretical, but unusually well-scaffolded and easy to enter.
-
- Homepage
- https://mit-mi.github.io/how2ai-course/spring2025/
- Topics
- Multimodal AI · Foundation models · Medical and sensory data · Audio and video
- Notes
- Graduate seminar on applying modern AI to unconventional data types. Less about any one application, more about the research mindset for tackling new modalities.
-
- Homepage
- https://www.scenerepresentations.org/courses/2025/spring/advances-in-cv/
- About
- Known for neural implicit representations (NeRF-adjacent work).
- Topics
- Neural scene representations · Multi-view geometry · Diffusion models · Contrastive learning · Embodied vision for robotics
- Notes
- Graduate course at the frontier of computer vision, from 3D scene understanding to generative models to robotic perception.
-
- Homepage
- https://cs50.harvard.edu/hls/2025/winter/
- About
- Built CS50 into one of the most visible publicly available computer-science course families.
- Topics
- Programming · Algorithms · SQL · Artificial intelligence · Web basics · Privacy and security
- Notes
- An accelerated January course for law students that explains technical systems in enough depth to reason about their legal consequences. More interdisciplinary than the usual CS50 spinoff, with daily assignments and a tight lecture sequence.
-
- Homepage
- https://ocw.mit.edu/courses/21h-151-dynastic-china-fall-2024/
- About
- MIT historian and the S. C. Fang Career Development Associate Professor of Chinese Language and Culture.
- Topics
- Imperial Chinese state formation · Chinese political thought · Dynastic transitions · Gender and social life · Commercial history · China in global context
- Notes
- A survey of Chinese history from the earliest dynasties to 1800, organized around state formation, intellectual life, commerce, and everyday society. The OCW release includes publicly watchable lecture videos, the full syllabus, and reading lists, making it a strong recent history addition to the catalog.
-
- Homepage
- https://ocw.mit.edu/courses/6-7960-deep-learning-fall-2024/
- About
- Isola is known for influential work in image-to-image translation and representation learning; Beery for applying ML to ecological monitoring; Bernstein for optimization theory in deep networks.
- Topics
- Neural network architectures · Learning theory · Backpropagation · Transformers · Geometry and invariances
- Notes
- A rigorous MIT graduate course on deep learning foundations, covering architecture families (CNNs, RNNs, graph nets, transformers) alongside approximation theory, generalization in high dimensions, and the geometry of learned representations. Unusually theory-forward for a deep learning course; fresh 2024 recording with full lecture notes and problem sets.
-
- Homepage
- https://ocw.mit.edu/courses/6-4590-foundations-of-information-policy-fall-2024/
- Topics
- Internet governance · Privacy · Cybersecurity · Freedom of expression · Intellectual property · AI policy
- Notes
- A policy-facing MIT course on how technical architecture and law shape the internet. The 2024 edition explicitly ties classic internet-policy debates to current AI questions, and the OCW release includes lecture notes, readings, and written assignments.
-
- Homepage
- https://ocw.mit.edu/courses/14-41-public-finance-and-public-policy-fall-2024/
- About
- Prominent public-finance and health-economics scholar; widely associated with the design of the Affordable Care Act.
- Topics
- Externalities · Public goods · Education policy · Health economics · Taxation · Social insurance
- Notes
- A full public-finance sequence with videos, handouts, problem sets, and solutions. Broad policy coverage rather than narrow technical specialization, which makes it a strong conceptual seed for economics on the page.
-
- Homepage
- https://cs50.harvard.edu/extension/ai/2024/fall/
- Topics
- Search · Knowledge representation · Probabilistic inference · Constraint satisfaction · Neural networks · Language
- Notes
- A compact survey of core AI ideas with one substantial project per unit. The material is broad rather than research-frontier, but it is a reliable open sequence that covers the classical conceptual map clearly.
-
- Homepage
- https://web.stanford.edu/class/cs234/CS234Spr2024/index.html
- About
- Known for work on safe RL, bandit algorithms, and efficient exploration; directs Stanford's AI safety and education research.
- Topics
- Markov decision processes · Policy gradients · Q-learning · Offline RL · Exploration · Value alignment
- Notes
- Stanford's main RL course, updated in 2024 with new content on DPO, offline RL, and LLM alignment. Includes a guest lecture on Direct Preference Optimization by the method's first authors. Full playlist publicly available on YouTube.
-
- Homepage
- https://deeplearning.cs.cmu.edu/S24/index.html
- About
- Raj and Singh are both Carnegie Mellon faculty known for speech, audio, and privacy-preserving machine learning.
- Topics
- MLPs · CNNs · RNNs · Attention mechanisms · Graph neural networks · Generative models
- Notes
- CMU's comprehensive deep learning sequence covering the full architecture family with thorough mathematical grounding. Unusually broad, with bootcamp labs on software foundations alongside lectures. 28 lectures publicly available on YouTube.
-
- Homepage
- https://ocw.mit.edu/courses/9-35-perception-spring-2024/
- About
- Leads MIT's Laboratory for Computational Audition; known for computational and psychophysical research on how humans perceive complex soundscapes.
- Topics
- Auditory perception · Visual system · Psychophysics · Color and motion perception · Object recognition · Chemical senses
- Notes
- A complete MIT undergraduate course covering the science of perception across the major senses, with emphasis on audition and vision. Uses illusion labs to probe perceptual mechanisms and applies psychophysical methods to quantify sensory thresholds and phenomena. The 23-lecture sequence was recorded in 2023–2024 and is freely available with full problem sets.
-
- Homepage
- https://web.stanford.edu/class/cs224n/
- About
- Co-author of the standard NLP textbooks and leads the Stanford NLP Group; known for foundational work in parsing, named entity recognition, and information extraction.
- Topics
- Word vectors · Transformers · Pre-training · Post-training · LLM agents · Benchmarking and reasoning
- Notes
- The canonical graduate NLP course at Stanford, updated for 2024 with new content on post-training, RLHF, reasoning, and agents. Lecture notes cover roughly the first half of the course. Full YouTube playlist is publicly available.
-
- Homepage
- https://rail.eecs.berkeley.edu/deeprlcourse/
- About
- Pioneer in deep RL, offline RL, and robot learning; leads Berkeley's Robotic AI and Learning Lab.
- Topics
- Imitation learning · Policy gradients · Actor-critic methods · Model-based RL · Inverse RL · Meta-learning
- Notes
- Berkeley's flagship deep RL course, known for rigorous mathematical treatment and research-frontier coverage. 23 lectures spanning classical policy optimization through offline RL, exploration, and meta-learning. The most recent publicly available full-semester recording.
-
- Homepage
- https://dlsyscourse.org/
- About
- Chen created TVM, XGBoost, and Apache MXNet; Kolter is known for implicit layers, equilibrium networks, and robust optimization.
- Topics
- Automatic differentiation · GPU computation · Neural network compilers · Operator fusion · Backpropagation implementation
- Notes
- A course about how deep learning frameworks actually work under the hood — students implement the Needle library from scratch, covering autodiff, CUDA ops, and compiler optimizations. A rare course that bridges ML and systems at a research depth. Individual 2022 lecture videos are publicly on YouTube.
-
- Homepage
- https://cs182sp21.github.io/
- Topics
- Backpropagation · CNNs · RNNs · Transformers · Meta-learning · Generative models
- Notes
- Berkeley's undergraduate deep learning course taught by Levine. Covers the full arc from backpropagation through meta-learning, with 21 lectures and 4 homework assignments. A solid public complement to CS285 for learners at the undergrad level.
-
- Homepage
- http://www.cs.cmu.edu/~odonnell/complexity17/
- About
- Known for Boolean function analysis and the textbook 'Analysis of Boolean Functions'; longstanding CMU theoretician.
- Topics
- Time and space hierarchy theorems · Circuit complexity · Randomized complexity · Interactive proofs · PCP theorem · Hardness amplification
- Notes
- A rigorous graduate course in computational complexity, covering the classical hierarchy from P and NP through circuit lower bounds, IP=PSPACE, and the PCP theorem. O'Donnell posts comprehensive handwritten lecture notes alongside 27 video lectures. One of the few fully public graduate complexity courses online.
-
- Homepage
- https://www.stevenstrogatz.com/teaching
- About
- Author of the canonical textbook on nonlinear dynamics and widely known for public writing on mathematics; based at Cornell Applied Mathematics.
- Topics
- Phase plane analysis · Bifurcations · Limit cycles · Lorenz equations · Chaos and strange attractors · Fractals
- Notes
- Strogatz's full 25-lecture graduate course filmed at Cornell, closely following his textbook. Covers 1D flows through chaos in the Lorenz system, with biological and physical applications throughout. Geometric and intuitive style that remains the gold standard introduction to the field.
-
- Homepage
- https://timroughgarden.org/f13/f13.html
- About
- Known for the price of anarchy in network routing and auction theory; now at Columbia after many years at Stanford.
- Topics
- Mechanism design · Vickrey and Myerson auctions · Price of anarchy · Selfish routing · No-regret learning · Nash equilibrium complexity
- Notes
- A graduate course at the intersection of algorithms and economics: auctions, mechanism design, equilibrium analysis, and the price of anarchy in networks. 20 video lectures plus comprehensive polished lecture notes. The standard reference for algorithmic game theory as a graduate topic.
-
- Homepage
- https://oyc.yale.edu/economics/econ-159
- About
- Professor of Economics and Management at Yale; known for his work on mechanism design and the history of economic thought.
- Topics
- Dominance · Nash equilibrium · Backward induction · Evolutionary stability · Asymmetric information · Auctions
- Notes
- One of the most celebrated open lecture series on game theory — 24 lectures that build from basic strategic reasoning to mechanism design, adverse selection, and signaling. Polak's teaching style is exceptionally clear and Socratic. Transcripts, problem sets, and video all free on Open Yale Courses.
-
- Homepage
- https://introtodeeplearning.com/
- About
- Alexander Amini leads MIT's deep autonomy research group; both instructors have run this annually-updated course since 2017, making it one of the most widely-viewed open MIT deep learning courses.
- Topics
- Deep learning fundamentals · Sequence modeling · Generative modeling · Reinforcement learning · Large language models · AI for science
- Notes
- MIT's annual short-course introduction to deep learning, updated for 2026 with new modules on AI for science, massively parallel training, and AI ethics. All lecture slides are open-sourced under MIT license and three practical labs are available on GitHub covering music generation, facial detection, and LLM fine-tuning.
-
- About
- Yale professor of German and Comparative Literature; co-editor of the new Princeton critical edition of Capital, Volume 1 (2024).
- Topics
- Political economy · Capital and labor · Value and commodities · Class struggle · Historical materialism · Accumulation
- Notes
- A chapter-by-chapter close reading of Capital Volume 1 taught by the co-editor of the 2024 Princeton critical edition. Nineteen lectures trace Marx's argument from commodity fetishism through the working day to primitive accumulation, with close attention to the philosophical architecture of the text.