NeurIPS 2019

Search Schedule »

Search:

Session:

Event type:

Hide breaks and receptions:

Day:

Show All »

Show All »

Dec. 8, 2019 Expo schedule » New in ML »

Timezone: America/Vancouver »

Indicates the beginning of a session

Time	West Ballroom A + B	West Exhibition Hall C + B3	West Exhibition Hall A	West Ballroom C
Tue Dec 10, 2019
10:05 AM (Orals)	Track 3 Session 1 Kernel Instrumental Variable Regression	Track 1 Session 1 Uniform convergence may be unable to explain generalization in deep learning	Track 2 Session 1 Logarithmic Regret for Online Control	Track 4 Session 1 Updates of Equilibrium Prop Match Gradients of Backprop Through Time in an RNN with Static Input
10:20 AM (Spotlights)	Machine Learning Estimation of Heterogeneous Treatment Effects with Instruments	On Exact Computation with an Infinitely Wide Neural Net	List-decodable Linear Regression	Legendre Memory Units: Continuous-Time Representation in Recurrent Neural Networks
10:25 AM (Spotlights)	Identification of Conditional Causal Effects under Markov Equivalence	Generalization Bounds of Stochastic Gradient Descent for Wide and Deep Neural Networks	On the Hardness of Robust Classification	Point-Voxel CNN for Efficient 3D Deep Learning
10:30 AM (Spotlights)	Likelihood-Free Overcomplete ICA and Applications In Causal Discovery	Efficient and Accurate Estimation of Lipschitz Constants for Deep Neural Networks	Adversarial Examples Are Not Bugs, They Are Features	Neural Networks with Cheap Differential Operators
10:35 AM (Spotlights)	Perceiving the arrow of time in autoregressive motion	Towards Explaining the Regularization Effect of Initial Large Learning Rate in Training Neural Networks	Empirically Measuring Concentration: Fundamental Limits on Intrinsic Robustness	Sequential Neural Processes
10:40 AM (Spotlights)	Conditional Independence Testing using Generative Adversarial Networks	Data-dependent Sample Complexity of Deep Neural Networks via Lipschitz Augmentation	Quantum Entropy Scoring for Fast Robust Mean Estimation and Improved Outlier Detection	Deep Equilibrium Models
04:10 PM (Orals)	Track 3 Session 2 Causal Confusion in Imitation Learning	Track 1 Session 2 Generative Modeling by Estimating Gradients of the Data Distribution	Track 2 Session 2 Trajectory of Alternating Direction Method of Multipliers and Adaptive Acceleration	Track 4 Session 2 Scalable Bayesian inference of dendritic voltage via spatiotemporal recurrent state space models
04:25 PM (Spotlights)	Imitation Learning from Observations by Minimizing Inverse Dynamics Disagreement	Learning Hierarchical Priors in VAEs	Asymmetric Valleys: Beyond Sharp and Flat Local Minima	Scalable Global Optimization via Local Bayesian Optimization
04:30 PM (Spotlights)	Learning to Control Self-Assembling Morphologies: A Study of Generalization via Modularity	Implicit Generation and Modeling with Energy Based Models	Reducing the variance in online optimization by transporting past gradients	Uncertainty on Asynchronous Time Event Prediction
04:35 PM (Spotlights)	A Structured Prediction Approach for Generalization in Cooperative Multi-Agent Reinforcement Learning	Invertible Convolutional Flow	Who is Afraid of Big Bad Minima? Analysis of gradient-flow in spiked matrix-tensor models	Bayesian Optimization under Heavy-tailed Payoffs
04:40 PM (Spotlights)	Learning Compositional Neural Programs with Recursive Tree Search and Planning	Residual Flows for Invertible Generative Modeling	Fast and Provable ADMM for Learning with Generative Priors	Variational Bayesian Optimal Experimental Design
04:45 PM (Spotlights)	Guided Meta-Policy Search	Dual Variational Generation for Low Shot Heterogeneous Face Recognition	Complexity of Highly Parallel Non-Smooth Convex Optimization	Implicit Posterior Variational Inference for Deep Gaussian Processes
04:50 PM (Orals)	Using a Logarithmic Mapping to Enable Lower Discount Factors in Reinforcement Learning	HYPE: A Benchmark for Human eYe Perceptual Evaluation of Generative Models	Necessary and Sufficient Geometries for Gradient Methods	Parameter elimination in particle Gibbs sampling
05:05 PM (Spotlights)	Better Exploration with Optimistic Actor Critic	Adaptive Density Estimation for Generative Models	SySCD: A System-Aware Parallel Coordinate Descent Algorithm	Divide and Couple: Using Monte Carlo Variational Objectives for Posterior Approximation
05:10 PM (Spotlights)	Robust exploration in linear quadratic reinforcement learning	Twin Auxilary Classifiers GAN	Learning Positive Functions with Pseudo Mirror Descent	The Randomized Midpoint Method for Log-Concave Sampling
05:15 PM (Spotlights)	Tight Regret Bounds for Model-Based Reinforcement Learning with Greedy Policies	Adversarial Fisher Vectors for Unsupervised Representation Learning	UniXGrad: A Universal, Adaptive Algorithm with Optimal Guarantees for Constrained Optimization	Poisson-Minibatching for Gibbs Sampling with Convergence Rate Guarantees
05:20 PM (Spotlights)	Hindsight Credit Assignment	Emergence of Object Segmentation in Perturbed Generative Models	Sinkhorn Barycenters with Free Support via Frank-Wolfe Algorithm	Stochastic Proximal Langevin Algorithm: Potential Splitting and Nonasymptotic Rates
05:25 PM (Spotlights)	Weight Agnostic Neural Networks	Compression with Flows via Local Bits-Back Coding	Learning dynamic polynomial proofs	Stochastic Runge-Kutta Accelerates Langevin Monte Carlo and Beyond

Time	West Ballroom A + B	West Exhibition Hall C + B3	West Exhibition Hall A	West Ballroom C
Wed Dec 11, 2019
10:05 AM (Orals)	Track 3 Session 3 Fast and Accurate Least-Mean-Squares Solvers	Track 1 Session 3 Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations	Track 2 Session 3 A neurally plausible model learns successor representations in partially observable environments	Track 4 Session 3 Faster width-dependent algorithm for mixed packing and covering LPs
10:20 AM (Spotlights)	Calibration tests in multi-class classification: A unifying framework	A Condition Number for Joint Optimization of Cycle-Consistent Networks	DualDICE: Behavior-Agnostic Estimation of Discounted Stationary Distribution Corrections	Sparse Logistic Regression Learns All Discrete Pairwise Graphical Models
10:25 AM (Spotlights)	Verified Uncertainty Calibration	Cascade RPN: Delving into High-Quality Region Proposal Network with Adaptive Convolution	VIREL: A Variational Inference Framework for Reinforcement Learning	Fast Convergence of Belief Propagation to Global Optima: Beyond Correlation Decay
10:30 AM (Spotlights)	Fast structure learning with modular regularization	Probabilistic Watershed: Sampling all spanning forests for seeded segmentation and semi-supervised learning	Unsupervised Curricula for Visual Meta-Reinforcement Learning	Smoothing Structured Decomposable Circuits
10:35 AM (Spotlights)	Principal Component Projection and Regression in Nearly Linear Time through Asymmetric SVRG	DM2C: Deep Mixed-Modal Clustering	Policy Continuation with Hindsight Inverse Dynamics	Counting the Optimal Solutions in Graphical Models
10:40 AM (Spotlights)	PIDForest: Anomaly Detection via Partial Identification	Learning Object Bounding Boxes for 3D Instance Segmentation on Point Clouds	Learning Reward Machines for Partially Observable Reinforcement Learning	Combining Generative and Discriminative Models for Hybrid Inference
03:50 PM (Orals)	Track 3 Session 4 Guided Similarity Separation for Image Retrieval	Track 1 Session 4 Dynamics of stochastic gradient descent for two-layer neural networks in the teacher-student setup	Track 2 Session 4 Batched Multi-armed Bandits Problem	Track 4 Session 4 Efficient and Thrifty Voting by Any Means Necessary
04:05 PM (Spotlights)	Fast and Flexible Multi-Task Classification using Conditional Neural Adaptive Processes	Time/Accuracy Tradeoffs for Learning a ReLU with respect to Gaussian Marginals	Are sample means in multi-armed bandits positively or negatively biased?	Finding Friend and Foe in Multi-Agent Games
04:10 PM (Spotlights)	Multimodal Model-Agnostic Meta-Learning via Task-Aware Modulation	Limitations of Lazy Training of Two-layers Neural Network	SIC-MMAB: Synchronisation Involves Communication in Multiplayer Multi-Armed Bandits	Efficient Regret Minimization Algorithm for Extensive-Form Correlated Equilibrium
04:15 PM (Spotlights)	Efficient Meta Learning via Minibatch Proximal Update	Generalization Bounds for Neural Networks via Approximate Description Length	Recovering Bandits	Poincaré Recurrence, Cycles and Spurious Equilibria in Gradient-Descent-Ascent for Non-Convex Non-Concave Zero-Sum Games
04:20 PM (Spotlights)	Reconciling meta-learning and continual learning with online mixtures of tasks	Small ReLU networks are powerful memorizers: a tight analysis of memorization capacity	Model Selection for Contextual Bandits	Multiagent Evaluation under Incomplete Information
04:25 PM (Orals)	Geometry-Aware Neural Rendering	On Making Stochastic Classifiers Deterministic	Strategizing against No-regret Learners	Distribution-Independent PAC Learning of Halfspaces with Massart Noise
04:40 PM (Spotlights)	Learning by Abstraction: The Neural State Machine	Cold Case: The Lost MNIST Digits	Beyond Online Balanced Descent: An Optimal Algorithm for Smoothed Online Optimization	Graph-based Discriminators: Sample Complexity and Expressiveness
04:45 PM (Spotlights)	Heterogeneous Graph Learning for Visual Commonsense Reasoning	An adaptive nearest neighbor rule for classification	Learning in Generalized Linear Contextual Bandits with Stochastic Delays	Nearly Tight Bounds for Robust Proper Learning of Halfspaces with a Margin
04:50 PM (Spotlights)	Self-Critical Reasoning for Robust Visual Question Answering	Multilabel reductions: what is my loss optimising?	Optimal Stochastic and Online Learning with Individual Iterates	McDiarmid-Type Inequalities for Graph-Dependent Variables and Stability Bounds
04:55 PM (Spotlights)	SuperGLUE: A Stickier Benchmark for General-Purpose Language Understanding Systems	Optimal Sparse Decision Trees	Online Learning via the Differential Privacy Lens	The Broad Optimality of Profile Maximum Likelihood

Time	West Ballroom A + B	West Exhibition Hall C + B3	West Exhibition Hall A	West Ballroom C
Thu Dec 12, 2019
10:05 AM (Orals)	Track 3 Session 5 Exponentially convergent stochastic k-PCA without variance reduction	Track 1 Session 5 Brain-Like Object Recognition with High-Performing Shallow Recurrent ANNs	Track 2 Session 5 Variance Reduction for Matrix Games	Track 4 Session 5 Optimizing Generalized Rate Metrics with Three Players
10:20 AM (Spotlights)	Implicit Regularization in Deep Matrix Factorization	Learning Perceptual Inference by Contrasting	Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers	Modeling Conceptual Understanding in Image Reference Games
10:25 AM (Spotlights)	SGD on Neural Networks Learns Functions of Increasing Complexity	Universality and individuality in neural dynamics across large populations of recurrent networks	Adversarial Music: Real world Audio Adversary against Wake-word Detection System	This Looks Like That: Deep Learning for Interpretable Image Recognition
10:30 AM (Spotlights)	When does label smoothing help?	Better Transfer Learning with Inferred Successor Maps	Convergence of Adversarial Training in Overparametrized Neural Networks	Assessing Social and Intersectional Biases in Contextualized Word Representations
10:35 AM (Spotlights)	Splitting Steepest Descent for Growing Neural Architectures	A unified theory for the origin of grid cells through the lens of pattern formation	Adversarial Training and Robustness for Multiple Perturbations	Paradoxes in Fair Machine Learning
10:40 AM (Spotlights)	Positional Normalization	Infra-slow brain dynamics as a marker for cognitive function and decline	Zero-shot Knowledge Transfer via Adversarial Belief Matching	Multi-Criteria Dimensionality Reduction with Applications to Fairness
03:50 PM (Orals)	Track 3 Session 6 Blind Super-Resolution Kernel Estimation using an Internal-GAN	Track 1 Session 6 Average Individual Fairness: Algorithms, Generalization and Experiments	Track 2 Session 6 Putting An End to End-to-End: Gradient-Isolated Learning of Representations	Track 4 Session 6 Understanding Sparse JL for Feature Hashing
04:05 PM (Spotlights)	Asymptotic Guarantees for Learning Generative Models with the Sliced-Wasserstein Distance	Ask not what AI can do, but what AI should do: Towards a framework of task delegability	On the Downstream Performance of Compressed Word Embeddings	A Nonconvex Approach for Exact and Efficient Multichannel Sparse Blind Deconvolution
04:10 PM (Spotlights)	Theoretical Analysis of Adversarial Learning: A Minimax Approach	Making AI Forget You: Data Deletion in Machine Learning	CPM-Nets: Cross Partial Multi-View Networks	Efficiently Learning Fourier Sparse Set Functions
04:15 PM (Spotlights)	Statistical bounds for entropic optimal transport: sample complexity and the central limit theorem	On Testing for Biases in Peer Review	Numerically Accurate Hyperbolic Embeddings Using Tiling-Based Models	Generalization Error Analysis of Quantized Compressive Learning
04:20 PM (Spotlights)	Differentiable Ranking and Sorting using Optimal Transport	A Step Toward Quantifying Independently Reproducible Machine Learning Research	Large Memory Layers with Product Keys	Surfing: Iterative Optimization Over Incrementally Trained Deep Networks
04:25 PM (Orals)	Nonparametric Density Estimation & Convergence Rates for GANs under Besov IPM Losses	On Robustness of Principal Component Regression	XLNet: Generalized Autoregressive Pretraining for Language Understanding	R2D2: Reliable and Repeatable Detector and Descriptor
04:40 PM (Spotlights)	KerGM: Kernelized Graph Matching	Private Learning Implies Online Learning: An Efficient Reduction	Cross-lingual Language Model Pretraining	Quadratic Video Interpolation
04:45 PM (Spotlights)	Wasserstein Weisfeiler-Lehman Graph Kernels	Private Stochastic Convex Optimization with Optimal Rates	N-Gram Graph: Simple Unsupervised Representation for Graphs, with Applications to Molecules	Training Image Estimators without Image Ground Truth
04:50 PM (Spotlights)	Regularization Matters: Generalization and Optimization of Neural Nets v.s. their Induced Kernel	Practical Differentially Private Top-k Selection with Pay-what-you-get Composition	Evaluating Protein Transfer Learning with TAPE	STREETS: A Novel Camera Network Dataset for Traffic Flow
04:55 PM (Spotlights)	Comparing distributions: $\ell_1$ geometry improves kernel two-sample testing	Differentially Private Markov Chain Monte Carlo	Cormorant: Covariant Molecular Neural Networks	Reflection Separation using a Pair of Unpolarized and Polarized Images