4/30/2019 |
Dalit Engelhardt |
Controlling stochastic evolution with deep reinforcement learning |
arxiv:1903.11373 |
Yes |
4/16/2019 |
|
Cancelled |
|
|
4/2/2019 |
|
Cancelled |
|
|
3/19/2019 |
Josh Batson |
Noise2Self: Blind Denoising by Self-Supervision |
arxiv:1901.11365 |
Yes |
3/5/2019 |
Boris Hanin |
Complexity of Linear Regions in Deep Networks |
arxiv:1901.09021 |
Yes |
2/19/2019 |
|
Cancelled |
|
|
2/5/2019 |
Jaehoon Lee |
Understanding wide neural networks |
arxiv:1711.00165 |
Yes |
1/22/2019 |
James Giammona |
Neural Ordinary Differential Equations |
arxiv:1806.07366 |
Yes |
1/8/2019 |
Yoni Kahn |
A Statistical Field Theory analogy |
arxiv:1710.06570 |
Yes |
12/25/2018 |
|
No Talk, Winter Break |
|
|
12/11/2018 |
Marat Freytsis |
Weak supervision, noisy labels, and error propagation |
arxiv:1706.09451, arxiv:1708.02949, arxiv:1402.5902 |
Yes |
11/27/2018 |
Stefano Spigler & Mario Geiger |
A jamming transition from under- to over-parametrization affects loss landscape and generalization |
arxiv:1809.09349, arxiv:1810.09665 |
Yes |
11/13/2018 |
Riccardo Zecchina |
The evolving entropy landscape of deep network |
Many |
|
10/30/2018 |
Marylou Gabrié |
Entropy and mutual information in models of deep neural networks |
arxiv:1805.09785 |
Yes |
10/16/2018 |
Sho Yaida |
Fluctuation-dissipation relations for SGD |
arxiv:1810.00004 |
Yes |
10/2/2018 |
Noam Brown |
Beating Humans at Poker |
science.aao1733 |
|
9/18/2018 |
Josh Batson |
Learning to Denoise Without Clean Data |
arXiv:1803.04189 |
Yes |
9/4/2018 |
Levent Sagun |
Continuation of the Exploration of the Loss Landscape: an Empirical View |
|
Yes |
8/21/2018 |
Boris Hanin |
Everything You Wanted to Know About the Loss Surface but Were Afraid to Ask – The Talk |
Many |
Yes |
8/7/2018 |
Yasaman Bahri |
Deep Learning and Quantum Entanglement |
arXiv:1704.01552 |
Yes |
7/24/2018 |
Adam Brown |
Fun in High-Dimensional Spaces |
|
|
7/10/2018 |
|
Cancelled due to ICML |
|
|
6/26/2018 |
David Schwab |
Renormalizing Data |
Koch-Janusz & Ringel and past work and a future idea |
Not Yet |
6/19/2018 |
Everyone |
Generative Query Networks – Discussion |
Eslami, et al. |
No |
6/12/2018 |
Ethan Dyer |
InfoGAN - Disentangled Latent Representations |
arXiv:1606.03657 |
Yes |
5/29/2018 |
Dima Krotov |
Dense Associative Memory & Adversarial Inputs |
NIPS 2016, arxiv:1701.00939 |
Yes |
5/15/2018 |
Eric Mintun |
Attention |
e.g. arxiv:1409.0473, arxiv:1702.00887, arxiv:1706.03762 |
Yes |
5/1/2018 |
Adam Brown |
A Critical Review of Quantum Machine Learning |
nature23474, arxiv:1611.09347 |
Yes |
4/17/2018 |
Josh Batson |
Equivariance in Deep Learning |
arxiv:1703.06114, arxiv:1801.10130 |
Yes |
4/3/2018 |
Shay Moran |
Generalization and Simplification in Statistical Learning |
|
Yes |
3/20/2018 |
Yoni Kahn |
Training Pruned Neural Networks |
arxiv:1803.03635 |
No |
3/6/2018 |
Paul Christiano |
AI Safety, Human Feedback, Robustness + Transparency |
arxiv:1706.03741 + others |
Yes |
2/20/2018 |
Jamie Sully |
Adversarial Examples |
arxiv:1801.02774 |
Yes |
2/6/2018 |
Boris Hanin |
When do exploding and vanishing gradients happen? |
arxiv:1801.03744 |
Yes |
1/23/2018 |
Guy Gur-Ari |
Three Factors Influencing Minima in SGD |
arxiv:1711.04623 |
Yes |
1/9/2018 |
Max Kleiman-Weiner |
Hierarchical RL and Eigenoptions |
arxiv:1703.00956, arxiv:1710.11089 |
Yes |
12/26/2017 |
Yoni Kahn |
Spontaneous Symmetry Breaking and Informtion Bottlenecks |
arxiv:1710.06096 |
Yes |
12/12/2017 |
Eric Mintun |
Capsule Networks |
Hinton, Krizhevsky, Wang, arxiv:1710.09829 |
Yes |
11/21/2017 |
Jared Kaplan |
Varitional Information Maximizing Exploration and Curiosity-Driven Self-Supervision |
arxiv:1605.09674, arxiv:1705.05363 |
Yes |
11/7/2017 |
DJ Strouse |
Information Bottlenecks and Variational Inference |
arxiv:1612.00410 |
No |
10/24/2017 |
Jamie Sully |
AlphaGo Zero |
nature24270 |
Yes |
10/10/2017 |
Dan Roberts |
Intro to RL and Distributional RL |
arxiv:1707.06887 |
Yes |
9/26/2017 |
Ethan Dyer |
Learning to RL & RL2 |
arxiv:1611.02779, arxiv:1611.05763 |
Yes |
9/12/2017 |
Guy Gur-Ari |
Learning to Learn and Metalearning |
arxiv:1606.04474, arxiv:1703.03400 |
Yes |
8/29/2017 |
Jaehoon Lee |
KFAC |
arxiv:1503.05671 |
No |