LeCun: 20 years of NYU deep learning teaching boils down to one framework — energy-based models
ylecun · x · 2026-11-21
Yann LeCun argues that his 20 years of teaching deep learning at NYU rest on one core idea: everything — classification, regression, structured prediction, and unsupervised/self-supervised methods like clustering, autoencoders, VAEs, Siamese nets, joint embedding, JEPA, and diffusion models — fits within the unified framework of energy-based models.
- Links the 2021 NYU DL course site with full lecture videos, slides and notebooks (Themes 3 & 4 cover EBM foundations and advanced topics)
- Notes @alfcnz is turning the course into an EBM textbook
- Also shares his 2019 IAS lecture on understanding SSL via EBMs and his 2006 EBM tutorial paper
More from Research
- Why does a Lean-verified proof get more hate than a contested journal proof? — IAmTimNguyen · 2026-11-21
- Mathematician: without Lean and Mathlib, the RH proof would have been a 250,000-page disaster — AlexKontorovich · 2026-11-21
- David Bessis: a 2M-line Lean Riemann proof would collapse the 'theorem economy' — beenwrekt · 2026-11-21
- Compute should go to better KV cache, not next-token decoding — YouJiacheng · 2026-11-21
- Nathan Lambert lays out six open problems in post-training in new Cornell slide deck — natolambert · 2026-11-21
- Fields Medal reportedly denied over AI-assisted Riemann proof, sparks debate — GiorgioPatrini · 2026-11-21