
Localized Calibration: Metrics and Recalibration
Probabilistic classifiers output confidence scores along with their pred...
read it

Neural Network Compression for Noisy Storage Devices
Compression and efficient storage of neural network (NN) parameters is c...
read it

Negative Data Augmentation
Data augmentation is often used to enlarge datasets with synthetic sampl...
read it

PrivacyConstrained Policies via Mutual Information Regularized Policy Gradients
As reinforcement learning techniques are increasingly applied to realwo...
read it

PiRank: Learning To Rank via Differentiable Sorting
A key challenge with machine learning approaches for ranking is the gap ...
read it

Efficient Conditional Pretraining for Transfer Learning
Almost all the stateoftheart neural networks for computer vision task...
read it

GeographyAware SelfSupervised Learning
Contrastive learning methods have significantly narrowed the gap between...
read it

Right Decisions from Wrong Predictions: A Mechanism Design Alternative to Individual Calibration
Decision makers often need to rely on imperfect probabilistic forecasts....
read it

Probabilistic Circuits for Variational Inference in Discrete Graphical Models
Inference in discrete graphical models with variational methods is diffi...
read it

Denoising Diffusion Implicit Models
Denoising diffusion probabilistic models (DDPMs) have achieved high qual...
read it

Using satellite imagery to understand and promote sustainable development
Accurate and comprehensive measurements of a range of sustainable develo...
read it

Privacy Preserving Recalibration under Domain Shift
Classifiers deployed in highstakes realworld applications must output ...
read it

HiPPO: Recurrent Memory with Optimal Polynomial Projections
A central problem in learning from sequential data is representing cumul...
read it

Multilabel Contrastive Predictive Coding
Variational mutual information (MI) estimators are widely used in unsupe...
read it

Belief Propagation Neural Networks
Learned neural solvers have successfully been used to solve combinatoria...
read it

Unsupervised Calibration under Covariate Shift
A probabilistic model is said to be calibrated if its predicted probabil...
read it

Experience Replay with Likelihoodfree Importance Weights
The use of past experiences to accelerate temporal difference (TD) learn...
read it

Individual Calibration with Randomized Forecasting
Machine learning applications often require calibrated predictions, e.g....
read it

A Framework for Sample Efficient Interval Estimation with Control Variates
We consider the problem of estimating confidence intervals for the mean ...
read it

Predicting Livelihood Indicators from Crowdsourced Street Level Images
Major decisions from governments and other large organizations rely on m...
read it

Efficient Poverty Mapping using Deep Reinforcement Learning
The combination of highresolution satellite imagery and machine learnin...
read it

Evaluating the Disentanglement of Deep Generative Models through Manifold Topology
Learning disentangled representations is regarded as a fundamental task ...
read it

MOPO: Modelbased Offline Policy Optimization
Offline reinforcement learning (RL) refers to the problem of learning po...
read it

Farmland Parcel Delineation Using Spatiotemporal Convolutional Networks
Farm parcel delineation provides cadastral data that is important in dev...
read it

Output Diversified Initialization for Adversarial Attacks
Adversarial examples are often constructed by iteratively refining a ran...
read it

Training Deep EnergyBased Models with fDivergence Minimization
Deep energybased models (EBMs) are very flexible in distribution parame...
read it

Gaussianization Flows
Iterative Gaussianization is a fixedpoint iteration procedure that can ...
read it

Predictive Coding for LocallyLinear Control
Highdimensional observations and unknown dynamics are major challenges ...
read it

Permutation Invariant Graph Generation via ScoreBased Generative Modeling
Learning generative models for graphstructured data is challenging beca...
read it

Learning When and Where to Zoom with Deep Reinforcement Learning
While high resolution images contain semantically more useful informatio...
read it

A Theory of Usable Information Under Computational Constraints
We propose a new framework for reasoning about information in complex sy...
read it

Nonlinear Equation Solving: A Faster Alternative to Feedforward Computation
Feedforward computations, such as evaluating a neural network or samplin...
read it

Generating Interpretable Poverty Maps using Object Detection in Satellite Images
Accurate locallevel poverty measurement is an essential task for govern...
read it

Cloud Removal in Satellite Images Using Spatiotemporal Generative Networks
Satellite images hold great promise for continuous environmental monitor...
read it

Efficient Object Detection in Large Images using Deep Reinforcement Learning
Traditionally, an object detector is applied to every part of the scene ...
read it

Approximating the Permanent by Sampling from Adaptive Partitions
Computing the permanent of a nonnegative matrix is a core problem with ...
read it

Fair Generative Modeling via Weak Supervision
Realworld datasets are often biased with respect to key demographic fac...
read it

Bridging the Gap Between fGANs and Wasserstein GANs
Generative adversarial networks (GANs) have enjoyed much success in lear...
read it

Weakly Supervised Disentanglement with Guarantees
Learning disentangled representations that correspond to factors of vari...
read it

Unsupervised OutofDistribution Detection with Batch Normalization
Likelihood from a generative model is a natural statistic for detecting ...
read it

Understanding the Limitations of Variational Mutual Information Estimators
Variational approaches based on neural networks are showing promise for ...
read it

Cross Domain Imitation Learning
We study the question of how to imitate tasks across domains with discre...
read it

MetaInverse Reinforcement Learning with Probabilistic Context Variables
Providing a suitable reward function to reinforcement learning can be di...
read it

Temporal FiLM: Capturing LongRange Sequence Dependencies with FeatureWise Modulations
Learning representations that accurately capture longrange dependencies...
read it

MultiAgent Adversarial Inverse Reinforcement Learning
Reinforcement learning agents are prone to undesired behaviors due to re...
read it

MintNet: Building Invertible Neural Networks with Masked Convolutions
We propose a new way of constructing invertible neural networks by combi...
read it

Generative Modeling by Estimating Gradients of the Data Distribution
We introduce a new generative model where samples are produced via Lange...
read it

Bias Correction of Learned Generative Models using LikelihoodFree Importance Weighting
A learned generative model often produces biased statistics relative to ...
read it

Calibrated ModelBased Deep Reinforcement Learning
Estimates of predictive uncertainty are important for accurate modelbas...
read it

Learning Neural PDE Solvers with Convergence Guarantees
Partial differential equations (PDEs) are widely used across the physica...
read it
Stefano Ermon
is this you? claim profile
Assistant Professor, Department of Computer Science Fellow, Woods Institute for the Environment at Stanford University