|
Interests
Machine Learning, Probabilistic Models, Optimization, Inference
Slides
Monte Carlo Variational Inference
Diffusion-based variational inference
Convergence Guarantees for Variational Inference
Selected Recent Publications
Hamiltonian Monte Carlo Inference of Marginalized Linear Mixed-Effects Models, NeurIPS 2024 (with J. Lai and D. Sheldon)
Disentangling impact of capacity, objective, batchsize, estimators, and step-size on flow VI, ICML 2024 (with A. Agrawal)
Understanding and mitigating difficulties in posterior predictive evaluation, ICML 2024 (with A. Agrawal)
Sample Average Approximation for Black-box VI, UAI 2024 (with J. Burroni and D. Sheldon)
Simulation-based stacking, AISTATS 2024 (with Y. Yao and B. Régaldo-Saint Blancard)
Joint control variate for faster black-box variational inference, AISTATS 2024 (with X. Wang and T. Geffner)
Provable convergence guarantees for black-box variational inference, NeurIPS 2023 (with G. Garrigos and R. Gower)
Discriminative calibration: Check Bayesian computation from simulations and flexible classifier, NeurIPS 2023 (with Y. Yao)
U-Statistics for Importance-Weighted Variational Inference, TMLR 2023 (with J. Burroni, K. Takatsu, D. Sheldon)
Langevin Diffusion Variational Inference, AISTATS 2022 (with T. Geffner)
Variational Inference with Locally Enhanced Bounds for Hierarchical Models, ICML 2022 (with T. Geffner)
Variational Marginal Particle Filters, AISTATS 2022 (with J. Lai and D. Sheldon)
Amortized Variational Inference for Simple Hierarchical Models, NeurIPS 2021 (with A. Agrawal)
MCMC Variational Inference via Uncorrected Hamiltonian Annealing, NeurIPS 2021 (with T. Geffner)
On the Difficulty of Unbiased Alpha Divergence Minimization, ICML 2021 (with T. Geffner) [talk] [poster] [slides]
Advances in Black-Box VI: Normalizing Flows, Importance Weighting, and Optimization, NeurIPS 2020 (with A. Agrawal and D. Sheldon) [code]
Approximation Based Variance Reduction for Reparameterization Gradients, NeurIPS 2020 (with T. Geffner)
Provable Smoothness Guarantees for Black-Box Variational Inference, ICML 2020 [talk] [slides]
A Rule for Gradient Estimator Selection, with an Application to Variational Inference, AISTATS 2020 (with T. Geffner) [talk]
Divide and Couple: Using Monte Carlo Variational Objectives for Posterior Approximation, NeurIPS 2019 (with D. Sheldon) [slides] [poster] [talk]
Provable Gradient Variance Guarantees for Black-Box Variational Inference, NeurIPS 2019 [poster]
Thompson Sampling and Approximate Inference, NeurIPS 2019 (with M. Phan and Y. Abbasi-Yadkori)
Using Large Ensembles of Control Variates for Variational Inference, NeurIPS 2018 (with T. Geffner)
Importance Weighting and Variational Inference, NeurIPS 2018 (with D. Sheldon) [poster]
| 
