Guest lectures (Fall 2019)

by @stats385 · last added 2026-09-24 · 9 items [lectures]

Each guest’s lecture page on the course site — speaker, date, and topic pointer for students.

─── 2026-09-24 ─────────────────────────────────────────

Guest lecture 12/4/19 — Vardan Papyan (Stanford). Hessians at scale, saddle points, and spectral views of deepnet training.

stats385.github.io · node›

Guest lecture 11/20/19 — Andrew Saxe (Oxford). Exact dynamics of deep linear networks and optimization geometry.

stats385.github.io · node›

Guest lecture 11/13/19 — Nati Srebro (TTI Chicago). Implicit bias of gradient methods and inductive bias in deep learning.

stats385.github.io · node›

Guest lecture 11/6/19 — Aleksander Madry (MIT). Adversarial robustness: attacks, defenses, and accuracy–robustness tradeoffs.

stats385.github.io · node›

Guest lecture 10/30/19 — Arthur Jacot (EPFL). Neural Tangent Kernel — convergence and generalization in the lazy regime.

stats385.github.io · node›

Guest lecture 10/23/19 — Song Mei (Stanford). Mean-field limits and landscape analysis for two-layer networks.

stats385.github.io · node›

Guest lecture 10/16/19 — Jeffrey Pennington (Google). Wide networks, Gaussian processes, and random-matrix views of deep learning.

stats385.github.io · node›

Guest lecture 10/9/19 — Tengyu Ma (Stanford). Optimization landscapes, generalization, and theory for deep nets.

stats385.github.io · node›

Guest lecture 10/2/19 — Stefano Soatto (UCLA). Invariance, information, and structure in deep representations.

stats385.github.io · node›

| all |