Sotabase

Career

· Assistant Professor, NYU Tandon School of Engineering2025–
· Researcher, Anthropic2024–
· Research Scientist, OpenAI2023–2024
· Research Intern, Google Brain2022–
· Research Intern / Student Researcher, Google2021–2022
· Research Intern, Google AI2020–
· Research Intern, Amazon AWS2019–
· PhD Student, New York University2019–2023
· PhD Student, Cornell University2017–2019
· Research Assistant, Lomonosov Moscow State University2013–2017

Publications (70)

Averaging Weights Leads to Wider Optima and Better Generalization
Conference on Uncertainty in Artificial Intelligence · 2018
1,903
cited
A Simple Baseline for Bayesian Uncertainty in Deep Learning
Neural Information Processing Systems · 2019
925
cited
Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs
Neural Information Processing Systems · 2018
856
cited
Bayesian Deep Learning and a Probabilistic Perspective of Generalization
Neural Information Processing Systems · 2020
746
cited
What Are Bayesian Neural Network Posteriors Really Like?
International Conference on Machine Learning · 2021
443
cited
Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations
International Conference on Learning Representations · 2022
431
cited
Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision
International Conference on Machine Learning · 2023
402
cited
Generalizing Convolutional Neural Networks for Equivariance to Lie Groups on Arbitrary Continuous Data
International Conference on Machine Learning · 2020
361
cited
Why Normalizing Flows Fail to Detect Out-of-Distribution Data
Neural Information Processing Systems · 2020
319
cited
Does Knowledge Distillation Really Work?
Neural Information Processing Systems · 2021
260
cited
There Are Many Consistent Explanations of Unlabeled Data: Why You Should Average
International Conference on Learning Representations · 2018
258
cited
On Feature Learning in the Presence of Spurious Correlations
Neural Information Processing Systems · 2022
172
cited
Subspace Inference for Bayesian Deep Learning
Conference on Uncertainty in Artificial Intelligence · 2019
155
cited
FlexiViT: One Model for All Patch Sizes
Computer Vision and Pattern Recognition · 2022
143
cited
Semi-Supervised Learning with Normalizing Flows
International Conference on Machine Learning · 2019
131
cited
Bayesian Model Selection, the Marginal Likelihood, and Generalization
International Conference on Machine Learning · 2022
77
cited
Learning Invariances in Neural Networks
Neural Information Processing Systems · 2020
77
cited
Tensor Train decomposition on TensorFlow (T3F)
Journal of machine learning research · 2018
77
cited
Simple and Fast Group Robustness by Automatic Feature Reweighting
International Conference on Machine Learning · 2023
75
cited
Scalable Gaussian Processes with Billions of Inducing Inputs via Tensor Train Decomposition
International Conference on Artificial Intelligence and Statistics · 2017
66
cited
Sotabase
Pavel Izmailov | Researcher Profile | Sotabase | Sotabase