【论文阅读笔记】NeurIPS2020文章列表Part1

• A graph similarity for deep learning
• An Unsupervised Information-Theoretic Perceptual Quality Metric
• Self-Supervised MultiModal Versatile Networks
• Benchmarking Deep Inverse Models over time, and the Neural-Adjoint method
• Off-Policy Evaluation and Learning for External Validity under a Covariate Shift
• Neural Methods for Point-wise Dependency Estimation
• Fast and Flexible Temporal Point Processes with Triangular Maps
• Backpropagating Linearly Improves Transferability of Adversarial Examples
• PyGlove: Symbolic Programming for Automated Machine Learning
• Fourier Sparse Leverage Scores and Approximate Kernel Learning
• Improved Algorithms for Online Submodular Maximization via First-order Regret Bounds
• Synbols: Probing Learning Algorithms with Synthetic Datasets
• Adversarially Robust Streaming Algorithms via Differential Privacy
• Trading Personalization for Accuracy: Data Debugging in Collaborative Filtering
• Cascaded Text Generation with Markov Transformers
• Improving Local Identifiability in Probabilistic Box Embeddings
• Permute-and-Flip: A new mechanism for differentially private selection
• Deep reconstruction of strange attractors from time series
• Reciprocal Adversarial Learning via Characteristic Functions
• Statistical Guarantees of Distributed Nearest Neighbor Classification
• Stein Self-Repulsive Dynamics: Benefits From Past Samples
• The Statistical Complexity of Early-Stopped Mirror Descent
• Algorithmic recourse under imperfect causal knowledge: a probabilistic approach
• Quantitative Propagation of Chaos for SGD in Wide Neural Networks
• A Causal View on Robustness of Neural Networks
• Minimax Classification with 0-1 Loss and Performance Guarantees
• How to Learn a Useful Critic? Model-based Action-Gradient-Estimator Policy Optimization
• Coresets for Regressions with Panel Data
• Learning Composable Energy Surrogates for PDE Order Reduction
• Efficient Contextual Bandits with Continuous Actions
• Achieving Equalized Odds by Resampling Sensitive Attributes
• Multi-Robot Collision Avoidance under Uncertainty with Probabilistic Safety Barrier Certificates
• Hard Shape-Constrained Kernel Machines
• A Closer Look at the Training Strategy for Modern Meta-Learning
• On the Value of Out-of-Distribution Testing: An Example of Goodhart’s Law
• Generalised Bayesian Filtering via Sequential Monte Carlo
• Deterministic Approximation for Submodular Maximization over a Matroid in Nearly Linear Time
• Flows for simultaneous manifold learning and density estimation
• Simultaneous Preference and Metric Learning from Paired Comparisons
• Efficient Variational Inference for Sparse Deep Learning with Theoretical Guarantee
• Learning Manifold Implicitly via Explicit Heat-Kernel Learning
• Deep Relational Topic Modeling via Graph Poisson Gamma Belief Network
• One-bit Supervision for Image Classification
• What is being transferred in transfer learning?
• Submodular Maximization Through Barrier Functions
• Neural Networks with Recurrent Generative Feedback
• Learning to Extrapolate Knowledge: Transductive Few-shot Out-of-Graph Link Prediction
• Exploiting weakly supervised visual patterns to learn from partial annotations
• Improving Inference for Neural Image Compression
• Neuron Merging: Compensating for Pruned Neurons
• FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
• Reinforcement Learning with Combinatorial Actions: An Application to Vehicle Routing
• Towards Playing Full MOBA Games with Deep Reinforcement Learning
• Rankmax: An Adaptive Projection Alternative to the Softmax Function
• Online Agnostic Boosting via Regret Minimization
• Causal Intervention for Weakly-Supervised Semantic Segmentation
• Belief Propagation Neural Networks
• Over-parameterized Adversarial Training: An Analysis Overcoming the Curse of Dimensionality
• Post-training Iterative Hierarchical Data Augmentation for Deep Networks
• Debugging Tests for Model Explanations
• Robust compressed sensing using generative models
• Fairness without Demographics through Adversarially Reweighted Learning
• Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Model
• Ridge Rider: Finding Diverse Solutions by Following Eigenvectors of the Hessian
• The route to chaos in routing games: When is price of anarchy too optimistic?
• Online Algorithm for Unsupervised Sequential Selection with Contextual Information
• Adapting Neural Architectures Between Domains
• What went wrong and when? Instance-wise feature importance for time-series black-box models
• Towards Better Generalization of Adaptive Gradient Methods
• Learning Guidance Rewards with Trajectory-space Smoothing
• Variance Reduction via Accelerated Dual Averaging for Finite-Sum Optimization
• Tree! I am no Tree! I am a low dimensional Hyperbolic Embedding
• Deep Structural Causal Models for Tractable Counterfactual Inference
• Convolutional Generation of Textured 3D Meshes
• A Statistical Framework for Low-bitwidth Training of Deep Neural Networks
• Better Set Representations For Relational Reasoning
• AutoSync: Learning to Synchronize for Data-Parallel Distributed Deep Learning
• A Combinatorial Perspective on Transfer Learning
• Hardness of Learning Neural Networks with Natural Weights
• Higher-Order Spectral Clustering of Directed Graphs
• Primal-Dual Mesh Convolutional Neural Networks
• The Advantage of Conditional Meta-Learning for Biased Regularization and Fine Tuning
• Watch out! Motion is Blurring the Vision of Your Deep Neural Networks
• Sinkhorn Barycenter via Functional Gradient Descent
• Coresets for Near-Convex Functions
• Bayesian Deep Ensembles via the Neural Tangent Kernel
• Improved Schemes for Episodic Memory-based Lifelong Learning
• Adaptive Sampling for Stochastic Risk-Averse Learning
• Deep Wiener Deconvolution: Wiener Meets Deep Learning for Image Deblurring
• Discovering Reinforcement Learning Algorithms
• Taming Discrete Integration via the Boon of Dimensionality
• Blind Video Temporal Consistency via Deep Video Prior
• Simplify and Robustify Negative Sampling for Implicit Collaborative Filtering
• Model Selection for Production System via Automated Online Experiments
• On the Almost Sure Convergence of Stochastic Gradient Descent in Non-Convex Problems
• Automatic Perturbation Analysis for Scalable Certified Robustness and Beyond
• Adaptation Properties Allow Identification of Optimized Neural Codes
• Global Convergence and Variance Reduction for a Class of Nonconvex-Nonconcave Minimax Problems
• Model-Based Multi-Agent RL in Zero-Sum Markov Games with Near-Optimal Sample Complexity
• Conservative Q-Learning for Offline Reinforcement Learning
• Online Influence Maximization under Linear Threshold Model
• Ensembling geophysical models with Bayesian Neural Networks
• Delving into the Cyclic Mechanism in Semi-supervised Video Object Segmentation
• Asymmetric Shapley values: incorporating causal knowledge into model-agnostic explainability
• Understanding Deep Architecture with Reasoning Layer
• Planning in Markov Decision Processes with Gap-Dependent Sample Complexity
• Provably Good Batch Off-Policy Reinforcement Learning Without Great Exploration
• Detection as Regression: Certified Object Detection with Median Smoothing
• Contextual Reserve Price Optimization in Auctions via Mixed Integer Programming
• ExpandNets: Linear Over-parameterization to Train Compact Convolutional Networks
• FleXOR: Trainable Fractional Quantization
• The Implications of Local Correlation on Learning Some Deep Functions
• Learning to search efficiently for causally near-optimal treatments
• A Game Theoretic Analysis of Additive Adversarial Attacks and Defenses
• Posterior Network: Uncertainty Estimation without OOD Samples via Density-Based Pseudo-Counts
• Recurrent Quantum Neural Networks
• No-Regret Learning and Mixed Nash Equilibria: They Do Not Mix
• A Unifying View of Optimism in Episodic Reinforcement Learning
• Continuous Submodular Maximization: Beyond DR-Submodularity
• An Asymptotically Optimal Primal-Dual Incremental Algorithm for Contextual Linear Bandits
• Assessing SATNet’s Ability to Solve the Symbol Grounding Problem
• A Bayesian Nonparametrics View into Deep Representations
• On the Similarity between the Laplace and Neural Tangent Kernels
• A causal view of compositional zero-shot recognition
• HiPPO: Recurrent Memory with Optimal Polynomial Projections
• Auto Learning Attention
• CASTLE: Regularization via Auxiliary Causal Graph Discovery
• Long-Tailed Classification by Keeping the Good and Removing the Bad Momentum Causal Effect
• Explainable Voting
• Deep Archimedean Copulas
• Re-Examining Linear Embeddings for High-Dimensional Bayesian Optimization
• UnModNet: Learning to Unwrap a Modulo Image for High Dynamic Range Imaging
• Thunder: a Fast Coordinate Selection Solver for Sparse Learning
• Neural Networks Fail to Learn Periodic Functions and How to Fix It
• Distribution Matching for Crowd Counting
• Correspondence learning via linearly-invariant embedding
• Learning to Dispatch for Job Shop Scheduling via Deep Reinforcement Learning
• On Adaptive Attacks to Adversarial Example Defenses
• Sinkhorn Natural Gradient for Generative Models
• Online Sinkhorn: Optimal Transport distances from sample streams
• Ultrahyperbolic Representation Learning
• Locally-Adaptive Nonparametric Online Learning
• Compositional Generalization via Neural-Symbolic Stack Machines
• Graphon Neural Networks and the Transferability of Graph Neural Networks
• Unreasonable Effectiveness of Greedy Algorithms in Multi-Armed Bandit with Many Arms
• Gamma-Models: Generative Temporal Difference Learning for Infinite-Horizon Prediction
• Deep Transformers with Latent Depth
• Neural Mesh Flow: 3D Manifold Mesh Generation via Diffeomorphic Flows
• Statistical control for spatio-temporal MEG/EEG source imaging with desparsified mutli-task Lasso
• A Scalable MIP-based Method for Learning Optimal Multivariate Decision Trees
• Efficient Exact Verification of Binarized Neural Networks
• Ultra-Low Precision 4-bit Training of Deep Neural Networks
• Bridging the Gap between Sample-based and One-shot Neural Architecture Search with BONAS
• On Numerosity of Deep Neural Networks
• Outlier Robust Mean Estimation with Subgaussian Rates via Stability
• Self-Supervised Relationship Probing
• Information Theoretic Counterfactual Learning from Missing-Not-At-Random Feedback
• Prophet Attention: Predicting Attention with Future Attention
• Language Models are Few-Shot Learners
• Margins are Insufficient for Explaining Gradient Boosting
• Fourier-transform-based attribution priors improve the interpretability and stability of deep learning models for genomics
• MomentumRNN: Integrating Momentum into Recurrent Neural Networks
• Marginal Utility for Planning in Continuous or Large Discrete Action Spaces
• Projected Stein Variational Gradient Descent
• Minimax Lower Bounds for Transfer Learning with Linear and One-hidden Layer Neural Networks
• SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks
• On the equivalence of molecular graph convolution and molecular wave function with poor basis set
• The Power of Predictions in Online Control
• Learning Affordance Landscapes for Interaction Exploration in 3D Environments
• Cooperative Multi-player Bandit Optimization
• Tight First- and Second-Order Regret Bounds for Adversarial Linear Bandits
• Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropout
• A Loss Function for Generative Neural Networks Based on Watson’s Perceptual Model
• Dynamic Fusion of Eye Movement Data and Verbal Narrations in Knowledge-rich Domains
• Scalable Multi-Agent Reinforcement Learning for Networked Systems with Average Reward
• Optimizing Neural Networks via Koopman Operator Theory
• SVGD as a kernelized Wasserstein gradient flow of the chi-squared divergence
• Adversarial Robustness of Supervised Sparse Coding
• Differentiable Meta-Learning of Bandit Policies
• Biologically Inspired Mechanisms for Adversarial Robustness
• Statistical-Query Lower Bounds via Functional Gradients
• Near-Optimal Reinforcement Learning with Self-Play
• Network Diffusions via Neural Mean-Field Dynamics
• Self-Distillation as Instance-Specific Label Smoothing
• Towards Problem-dependent Optimal Learning Rates
• Cross-lingual Retrieval for Iterative Self-Supervised Training
• Rethinking pooling in graph neural networks
• Pointer Graph Networks
• Gradient Regularized V-Learning for Dynamic Treatment Regimes
• Faster Wasserstein Distance Estimation with the Sinkhorn Divergence
• Forethought and Hindsight in Credit Assignment
• Robust Recursive Partitioning for Heterogeneous Treatment Effects with Uncertainty Quantification
• Rescuing neural spike train models from bad MLE
• Lower Bounds and Optimal Algorithms for Personalized Federated Learning
• Black-Box Certification with Randomized Smoothing: A Functional Optimization Based Framework
• Deep Imitation Learning for Bimanual Robotic Manipulation
• Stationary Activations for Uncertainty Calibration in Deep Learning
• Ensemble Distillation for Robust Model Fusion in Federated Learning
• Falcon: Fast Spectral Inference on Encrypted Data
• On Power Laws in Deep Ensembles
• Practical Quasi-Newton Methods for Training Deep Neural Networks
• Approximation Based Variance Reduction for Reparameterization Gradients
• Inference Stage Optimization for Cross-scenario 3D Human Pose Estimation
• Consistent feature selection for analytic deep neural networks
• Glance and Focus: a Dynamic Approach to Reducing Spatial Redundancy in Image Classification
• Information Maximization for Few-Shot Learning
• Inverse Reinforcement Learning from a Gradient-based Learner
• Bayesian Multi-type Mean Field Multi-agent Imitation Learning
• Bayesian Robust Optimization for Imitation Learning
• Multiview Neural Surface Reconstruction by Disentangling Geometry and Appearance
• Riemannian Continuous Normalizing Flows
• Attention-Gated Brain Propagation: How the brain can implement reward-based error backpropagation
• Asymptotic Guarantees for Generative Modeling Based on the Smooth Wasserstein Distance
• Online Robust Regression via SGD on the l1 loss
• PRANK: motion Prediction based on RANKing
• Fighting Copycat Agents in Behavioral Cloning from Observation Histories
• Tight Nonparametric Convergence Rates for Stochastic Gradient Descent under the Noiseless Linear Model
• Structured Prediction for Conditional Meta-Learning
• Optimal Lottery Tickets via Subset Sum: Logarithmic Over-Parameterization is Sufficient
• The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memes
• Stochasticity of Deterministic Gradient Descent: Large Learning Rate for Multiscale Objective Function
• Identifying Learning Rules From Neural Network Observables
• Optimal Approximation - Smoothness Tradeoffs for Soft-Max Functions
• Weakly-Supervised Reinforcement Learning for Controllable Behavior
• Improving Policy-Constrained Kidney Exchange via Pre-Screening
• Learning abstract structure for drawing by efficient motor program induction
• Why Do Deep Residual Networks Generalize Better than Deep Feedforward Networks? — A Neural Tangent Kernel Perspective
• Dual Instrumental Variable Regression
• Stochastic Gradient Descent in Correlated Settings: A Study on Gaussian Processes
• Interventional Few-Shot Learning
• Minimax Value Interval for Off-Policy Evaluation and Policy Optimization
• Biased Stochastic First-Order Methods for Conditional Stochastic Optimization and Applications in Meta Learning
• ShiftAddNet: A Hardware-Inspired Deep Network
• Network-to-Network Translation with Conditional Invertible Neural Networks
• Intra-Processing Methods for Debiasing Neural Networks
• Finding Second-Order Stationary Points Efficiently in Smooth Nonconvex Linearly Constrained Optimization Problems
• Model-based Policy Optimization with Unsupervised Model Adaptation
• Implicit Regularization and Convergence for Weight Normalization
• Geometric All-way Boolean Tensor Decomposition
• Modular Meta-Learning with Shrinkage
• A/B Testing in Dense Large-Scale Networks: Design and Inference
• What Neural Networks Memorize and Why: Discovering the Long Tail via Influence Estimation
• Partially View-aligned Clustering
• Partial Optimal Tranport with applications on Positive-Unlabeled Learning
• Toward the Fundamental Limits of Imitation Learning
• Logarithmic Pruning is All You Need
• Hold me tight! Influence of discriminative features on deep network boundaries
• Learning from Mixtures of Private and Public Populations
• Adversarial Weight Perturbation Helps Robust Generalization
• Stateful Posted Pricing with Vanishing Regret via Dynamic Deterministic Markov Decision Processes
• Adversarial Self-Supervised Contrastive Learning
• Normalizing Kalman Filters for Multivariate Time Series Analysis
• Learning to summarize with human feedback
• Fourier Spectrum Discrepancies in Deep Network Generated Images
• Lamina-specific neuronal properties promote robust, stable signal propagation in feedforward networks
• Learning Dynamic Belief Graphs to Generalize on Text-Based Games
• Triple descent and the two kinds of overfitting: where & why do they appear?
• Multimodal Graph Networks for Compositional Generalization in Visual Question Answering
• Learning Graph Structure With A Finite-State Automaton Layer
• A Universal Approximation Theorem of Deep Neural Networks for Expressing Probability Distributions
• Unsupervised object-centric video generation and decomposition in 3D
• Domain Generalization for Medical Imaging Classification with Linear-Dependency Regularization
• Multi-label classification: do Hamming loss and subset accuracy really conflict with each other?
• A Novel Automated Curriculum Strategy to Solve Hard Sokoban Planning Instances
• Causal analysis of Covid-19 Spread in Germany
• Locally private non-asymptotic testing of discrete distributions is faster using interactive mechanisms
• Adaptive Gradient Quantization for Data-Parallel SGD
• Finite Continuum-Armed Bandits
• Removing Bias in Multi-modal Classifiers: Regularization by Maximizing Functional Entropies
• Compact task representations as a normative model for higher-order brain activity
• Robust-Adaptive Control of Linear Systems: beyond Quadratic Costs
• Co-exposure Maximization in Online Social Networks
• UCLID-Net: Single View Reconstruction in Object Space
• Reinforcement Learning for Control with Multiple Frequencies
• Complex Dynamics in Simple Neural Networks: Understanding Gradient Flow in Phase Retrieval
• Neural Message Passing for Multi-Relational Ordered and Recursive Hypergraphs
• A Unified View of Label Shift Estimation
• Optimal Private Median Estimation under Minimal Distributional Assumptions
• Breaking the Communication-Privacy-Accuracy Trilemma
• Audeo: Audio Generation for a Silent Performance Video
• Ode to an ODE
• Self-Distillation Amplifies Regularization in Hilbert Space
• Coupling-based Invertible Neural Networks Are Universal Diffeomorphism Approximators
• Community detection using fast low-cardinality semidefinite programming

• Modeling Noisy Annotations for Crowd Counting
• An operator view of policy gradient methods
• Demystifying Contrastive Self-Supervised Learning: Invariances, Augmentations and Dataset Biases
• Online MAP Inference of Determinantal Point Processes
• Video Object Segmentation with Adaptive Feature Bank and Uncertain-Region Refinement
• Inferring learning rules from animal decision-making
• Input-Aware Dynamic Backdoor Attack
• How hard is to distinguish graphs with graph neural networks?
• Minimax Regret of Switching-Constrained Online Convex Optimization: No Phase Transition
• Dual Manifold Adversarial Robustness: Defense against Lp and non-Lp Adversarial Attacks
• Cross-Scale Internal Graph Neural Network for Image Super-Resolution
• Unsupervised Representation Learning by Invariance Propagation
• Restoring Negative Information in Few-Shot Object Detection
• Do Adversarially Robust ImageNet Models Transfer Better?
• Robust Correction of Sampling Bias using Cumulative Distribution Functions
• Personalized Federated Learning with Theoretical Guarantees: A Model-Agnostic Meta-Learning Approach
• Pixel-Level Cycle Association: A New Perspective for Domain Adaptive Semantic Segmentation
• Classification with Valid and Adaptive Coverage
• Learning Global Transparent Models consistent with Local Contrastive Explanations
• Learning to Approximate a Bregman Divergence
• Diverse Image Captioning with Context-Object Split Latent Spaces
• Le