Search Results for author:
Found 59 papers, 39 papers with code
Experts Don't Cheat: Learning What You Don't Know By Predicting Pairs
Identifying how much a model ${\widehat{p}}_{\theta}(Y|X)$ knows about the stochastic real-world process $p(Y|X)$ it was trained on is important to ensure it avoids producing incorrect or "hallucinated" answers or taking unsafe actions.
Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
We find that such backdoor behavior can be made persistent, so that it is not removed by standard safety training techniques, including supervised fine-tuning, reinforcement learning, and adversarial training (eliciting unsafe behavior and then training to remove it).
Sorting Out Quantum Monte Carlo
Molecular modeling at the quantum level requires choosing a parameterization of the wavefunction that both respects the required particle symmetries, and is scalable to systems of many particles.
Towards Understanding Sycophancy in Language Models
Overall, our results indicate that sycophancy is a general behavior of state-of-the-art AI assistants, likely driven in part by human preference judgments favoring sycophantic responses.
On Implicit Bias in Overparameterized Bilevel Optimization
Bilevel problems consist of two nested sub-problems, called the outer and inner problems, respectively.
Meta-Learning to Improve Pre-Training
Pre-training (PT) followed by fine-tuning (FT) is an effective method for training neural networks, and has led to significant performance improvements in many domains.
Complex Momentum for Optimization in Games
We generalize gradient descent with momentum for optimization in differentiable games to have complex-valued momentum.
Infinitely Deep Bayesian Neural Networks with Stochastic Differential Equations
We perform scalable approximate inference in continuous-depth Bayesian neural networks.
Oops I Took A Gradient: Scalable Sampling for Discrete Distributions
We propose a general and scalable approximate sampling strategy for probabilistic models with discrete variables.
Self-Tuning Stochastic Optimization with Curvature-Aware Gradient Filtering
Standard first-order stochastic optimization algorithms base their updates solely on the average mini-batch gradient, and it has been shown that tracking additional quantities such as the curvature can help de-sensitize common hyperparameters.
Teaching with Commentaries
We find that commentaries can improve training speed and/or performance, and provide insights about the dataset and training process.
No MCMC for me: Amortized sampling for fast and stable training of energy-based models
Energy-Based Models (EBMs) present a flexible and appealing way to represent uncertainty.
Learning Differential Equations that are Easy to Solve
Differential equations parameterized by neural networks become expensive to solve numerically as training progresses.
A Study of Gradient Variance in Deep Learning
We introduce a method, Gradient Clustering, to minimize the variance of average mini-batch gradient with stratified sampling.
SUMO: Unbiased Estimation of Log Marginal Probability for Latent Variable Models
Standard variational lower bounds used to train latent variable models produce biased estimates of most quantities of interest.
What went wrong and when? Instance-wise Feature Importance for Time-series Models
Explanations of time series models are useful for high stakes applications like healthcare but have received little attention in machine learning literature.
Learning the Stein Discrepancy for Training and Evaluating Energy-Based Models without Sampling
We estimate the Stein discrepancy between the data density $p(x)$ and the model density $q(x)$ defined by a vector function of the data.
Scalable Gradients for Stochastic Differential Equations
The adjoint sensitivity method scalably computes gradients of solutions to ordinary differential equations.
Ranked #1 on
Video Prediction
on CMU Mocap-2
Neural Networks with Cheap Differential Operators
Gradients of neural networks can be computed efficiently for any architecture, but some applications require differential operators with higher time complexity.
Your Classifier is Secretly an Energy Based Model and You Should Treat it Like One
In this setting, the standard class probabilities can be easily computed as well as unnormalized values of p(x) and p(x|y).
Optimizing Millions of Hyperparameters by Implicit Differentiation
We propose an algorithm for inexpensive gradient-based hyperparameter optimization that combines the implicit function theorem (IFT) with efficient inverse Hessian approximations.
Efficient Graph Generation with Graph Recurrent Attention Networks
Our model generates graphs one block of nodes and associated edges at a time.
Explaining Time Series by Counterfactuals
We propose a method to automatically compute the importance of features at every observation in time series, by simulating counterfactual trajectories given previous observations.
Understanding Undesirable Word Embedding Associations
Word embeddings are often criticized for capturing undesirable word associations such as gender stereotypes.
Latent ODEs for Irregularly-Sampled Time Series
Time series with non-uniform intervals occur in many applications, and are difficult to model using standard recurrent neural networks (RNNs).
Ranked #1 on
Multivariate Time Series Imputation
on MuJoCo
Multivariate Time Series Forecasting
Multivariate Time Series Imputation
+3
Residual Flows for Invertible Generative Modeling
Flow-based generative models parameterize probability distributions through an invertible transformation and can be trained by maximum likelihood.
Ranked #2 on
Image Generation
on MNIST
Self-Tuning Networks: Bilevel Optimization of Hyperparameters using Structured Best-Response Functions
Empirically, our approach outperforms competing hyperparameter optimization methods on large-scale deep learning problems.
Invertible Residual Networks
We show that standard ResNet architectures can be made invertible, allowing the same model to be used for classification, density estimation, and generation.
Ranked #5 on
Image Generation
on MNIST
Scalable Recommender Systemsthrough Recursive Evidence Chains
A popular matrix completion algorithm is matrix factorization, where ratings are predicted from combining learned user and item parameter vectors.
Towards Understanding Linear Word Analogies
A surprising property of word vectors is that word analogies can often be solved with vector arithmetic.
FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Models
The result is a continuous-time invertible generative model with unbiased density estimation and one-pass sampling, while allowing unrestricted neural network architectures.
Ranked #1 on
Density Estimation
on UCI MINIBOONE
Stochastic Combinatorial Ensembles for Defending Against Adversarial Examples
Many deep learning algorithms can be easily fooled with simple adversarial examples.
Explaining Image Classifiers by Counterfactual Generation
We can rephrase this question to ask: which parts of the image, if they were not seen by the classifier, would most change its decision?
Scalable Recommender Systems through Recursive Evidence Chains
Recommender systems can be formulated as a matrix completion problem, predicting ratings from user and item parameter vectors.
Neural Ordinary Differential Equations
Instead of specifying a discrete sequence of hidden layers, we parameterize the derivative of the hidden state using a neural network.
Ranked #2 on
Multivariate Time Series Imputation
on MuJoCo
Multivariate Time Series Forecasting
Multivariate Time Series Imputation
Stochastic Hyperparameter Optimization through Hypernetworks
Machine learning models are often tuned by nesting optimization of model weights inside the optimization of hyperparameters.
Isolating Sources of Disentanglement in Variational Autoencoders
We decompose the evidence lower bound to show the existence of a term measuring the total correlation between latent variables.
Inference Suboptimality in Variational Autoencoders
Furthermore, we show that the parameters used to increase the expressiveness of the approximation play a role in generalizing inference rather than simply improving the complexity of the approximation.
Generating and designing DNA with deep generative models
We propose generative neural network methods to generate DNA sequences and tune them to have desired properties.
Noisy Natural Gradient as Variational Inference
Variational Bayesian neural nets combine the flexibility of deep learning with Bayesian uncertainty estimation.
Backpropagation through the Void: Optimizing control variates for black-box gradient estimation
Gradient-based optimization is the foundation of deep learning and reinforcement learning.
Reinterpreting Importance-Weighted Autoencoders
The standard interpretation of importance-weighted autoencoders is that they maximize a tighter lower bound on the marginal likelihood than the standard evidence lower bound.
Sticking the Landing: Simple, Lower-Variance Gradient Estimators for Variational Inference
We propose a simple and general variant of the standard reparameterized gradient estimator for the variational evidence lower bound.
Automatic chemical design using a data-driven continuous representation of molecules
We report a method to convert discrete representations of molecules to and from a multidimensional continuous representation.
Neural networks for the prediction organic chemistry reactions
Reaction prediction remains one of the major challenges for organic chemistry, and is a pre-requisite for efficient synthetic planning.
Composing graphical models with neural networks for structured representations and fast inference
We propose a general modeling and inference framework that composes probabilistic graphical models with deep learning methods and combines their respective strengths.
Convolutional Networks on Graphs for Learning Molecular Fingerprints
We introduce a convolutional neural network that operates directly on graphs.
Ranked #2 on
Drug Discovery
on HIV dataset
Early Stopping is Nonparametric Variational Inference
By tracking the change in entropy over this sequence of transformations during optimization, we form a scalable, unbiased estimate of the variational lower bound on the log marginal likelihood.
Gradient-based Hyperparameter Optimization through Reversible Learning
Tuning hyperparameters of learning algorithms is hard because gradients are usually unavailable.
Raiders of the Lost Architecture: Kernels for Bayesian Optimization in Conditional Parameter Spaces
In practical Bayesian optimization, we must often search over structures with differing numbers of parameters.
Warped Mixtures for Nonparametric Cluster Shapes
A mixture of Gaussians fit to a single curved or heavy-tailed cluster will report that the data contains many clusters.
Optimally-Weighted Herding is Bayesian Quadrature
We show that the criterion minimised when selecting samples in kernel herding is equivalent to the posterior variance in Bayesian quadrature.
Probabilistic ODE Solvers with Runge-Kutta Means
We construct a family of probabilistic numerical methods that instead return a Gauss-Markov process defining a probability distribution over the ODE solution.
Avoiding pathologies in very deep networks
Choosing appropriate architectures and regularization strategies for deep networks is crucial to good predictive performance.
Automatic Construction and Natural-Language Description of Nonparametric Regression Models
This paper presents the beginnings of an automatic statistician, focusing on regression problems.
Structure Discovery in Nonparametric Regression through Compositional Kernel Search
Despite its importance, choosing the structural form of the kernel in nonparametric regression remains a black art.
Warped Mixtures for Nonparametric Cluster Shapes
A mixture of Gaussians fit to a single curved or heavy-tailed cluster will report that the data contains many clusters.
Optimally-Weighted Herding is Bayesian Quadrature
We show that the criterion minimised when selecting samples in kernel herding is equivalent to the posterior variance in Bayesian quadrature.
