Lecture preview

Student Sign-In Required

Full-length MP4 lecture streaming requires a free SeekSky AI Academy account to prevent automated bandwidth scraping.

MIT 15.773 · Lecture 01 57m 04s · 103.8 MB · CC BY-NC-SA 4.0

1. Introduction to Neural Networks & Deep Learning

Instructor: Prof. Rama Ramakrishnan · MIT Sloan School of Management

Lecture Key Concepts

    Open AI & Deep Learning Video Library

    Three complete university lecture series spanning foundational ML linear algebra, applied Hugging Face & RAG engineering, and LLM scaling theory.

    Structured Learning Track

    Track 01 · MIT 18.065

    Matrix Calculus & ML Math

    Build geometric intuition for column spaces, Singular Value Decomposition (SVD), and randomized matrix multiplication behind LoRA and embeddings.

    Track 02 · MIT 15.773

    Applied Deep Learning & LLMs

    Hands-on progression from perceptrons, CNNs, and Hugging Face transfer learning to Transformers, RAG pipelines, LoRA fine-tuning, and Diffusion.

    Track 03 · MIT 6.7960

    Scaling Laws & Frontier Architectures

    Explore approximation theory, optimization scaling rules, Transformer internals, generative modeling theory, and compute-optimal LLM scaling laws.

    Platform Information

    Why is registration required to stream videos?

    Full-length university lecture videos range from 65 MB to 215 MB each. Requiring free student authentication prevents unauthenticated web scrapers and bots from draining origin bandwidth while allowing members to save playback progress.

    Are these courses free and openly licensed?

    Yes. All 21 lectures are redistributed unmodified from MIT OpenCourseWare (Courses 15.773 Spring 2024, 6.7960 Fall 2024, and 18.065 Spring 2018) under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 license.

    What playback format is used?

    Videos are encoded as fast-start progressive H.264 MP4 (360p 16:9) supporting HTTP/2 byte-range seeking and instant resume across desktop and mobile browsers.