Sewoong Oh

papers

1,526

total citations

papers (25)

Quality Not Quantity: On the Interaction between Dataset Design and Robustness of CLIP

NEURIPS 2022arXiv

122

citations

Robust and differentially private mean estimation

NEURIPS 2021arXiv

citations

Label Poisoning is All You Need

NEURIPS 2023arXiv

citations

Learning To Generate Image Embeddings With User-Level Differential Privacy

CVPR 2023arXiv

citations

Robust Meta-learning for Mixed Linear Regression with Small Batches

NEURIPS 2020arXiv

citations

Privacy-Preserving Instructions for Aligning Large Language Models

ICML 2024arXiv

citations

On the Connection between Pre-training Data Diversity and Fine-tuning Robustness

NEURIPS 2023arXiv

citations

DP-PCA: Statistically Optimal and Differentially Private PCA

NEURIPS 2022arXiv

citations

Unleashing the Power of Randomization in Auditing Differentially Private ML

NEURIPS 2023arXiv

citations

Projection Efficient Subgradient Method and Optimal Nonsmooth Frank-Wolfe Method

NEURIPS 2020arXiv

citations

DPZero: Private Fine-Tuning of Language Models without Backpropagation

ICML 2024arXiv

citations

Bring Your Own Algorithm for Optimal Differentially Private Stochastic Minimax Optimization

NEURIPS 2022arXiv

citations

Divergence Frontiers for Generative Models: Sample Complexity, Quantization Effects, and Frontier Integrals

NEURIPS 2021arXiv

citations

Private (Stochastic) Non-Convex Optimization Revisited: Second-Order Stationary Points and Excess Risks

NEURIPS 2023arXiv

citations

DeepPolar: Inventing Nonlinear Large-Kernel Polar Codes via Deep Learning

ICML 2024arXiv

citations

PLeaS - Merging Models with Permutations and Least Squares

CVPR 2025arXiv

citations

Scalable Fingerprinting of Large Language Models

NEURIPS 2025arXiv

citations

Improved Communication-Privacy Trade-offs in $L_2$ Mean Estimation under Streaming Differential Privacy

ICML 2024arXiv

citations

Zonotope Domains for Lagrangian Neural Network Verification

NEURIPS 2022arXiv

citations

Understanding the Gain from Data Filtering in Multimodal Contrastive Learning

NEURIPS 2025arXiv

citations

Label Robust and Differentially Private Linear Regression: Computational and Statistical Efficiency

NEURIPS 2023

citations

Statistically and Computationally Efficient Linear Meta-representation Learning

NEURIPS 2021

citations

Sewoong Oh

papers (25)

DataComp: In search of the next generation of multimodal datasets

Gradient Inversion with Generative Image Prior

Improving multimodal datasets with image captioning

Quality Not Quantity: On the Interaction between Dataset Design and Robustness of CLIP

Robust and differentially private mean estimation

Label Poisoning is All You Need

Learning To Generate Image Embeddings With User-Level Differential Privacy

Robust Meta-learning for Mixed Linear Regression with Small Batches

Privacy-Preserving Instructions for Aligning Large Language Models

On the Connection between Pre-training Data Diversity and Fine-tuning Robustness

DP-PCA: Statistically Optimal and Differentially Private PCA

Unleashing the Power of Randomization in Auditing Differentially Private ML

Projection Efficient Subgradient Method and Optimal Nonsmooth Frank-Wolfe Method

DPZero: Private Fine-Tuning of Language Models without Backpropagation

Bring Your Own Algorithm for Optimal Differentially Private Stochastic Minimax Optimization

Divergence Frontiers for Generative Models: Sample Complexity, Quantization Effects, and Frontier Integrals

Private (Stochastic) Non-Convex Optimization Revisited: Second-Order Stationary Points and Excess Risks

DeepPolar: Inventing Nonlinear Large-Kernel Polar Codes via Deep Learning

PLeaS - Merging Models with Permutations and Least Squares

Scalable Fingerprinting of Large Language Models

Improved Communication-Privacy Trade-offs in $L_2$ Mean Estimation under Streaming Differential Privacy

Zonotope Domains for Lagrangian Neural Network Verification

Understanding the Gain from Data Filtering in Multimodal Contrastive Learning

Label Robust and Differentially Private Linear Regression: Computational and Statistical Efficiency

Statistically and Computationally Efficient Linear Meta-representation Learning

papers (25)

DataComp: In search of the next generation of multimodal datasets

Gradient Inversion with Generative Image Prior

Improving multimodal datasets with image captioning

Quality Not Quantity: On the Interaction between Dataset Design and Robustness of CLIP

Robust and differentially private mean estimation

Label Poisoning is All You Need

Learning To Generate Image Embeddings With User-Level Differential Privacy

Robust Meta-learning for Mixed Linear Regression with Small Batches

Privacy-Preserving Instructions for Aligning Large Language Models

On the Connection between Pre-training Data Diversity and Fine-tuning Robustness

DP-PCA: Statistically Optimal and Differentially Private PCA

Unleashing the Power of Randomization in Auditing Differentially Private ML

Projection Efficient Subgradient Method and Optimal Nonsmooth Frank-Wolfe Method

DPZero: Private Fine-Tuning of Language Models without Backpropagation

Bring Your Own Algorithm for Optimal Differentially Private Stochastic Minimax Optimization

Divergence Frontiers for Generative Models: Sample Complexity, Quantization Effects, and Frontier Integrals

Private (Stochastic) Non-Convex Optimization Revisited: Second-Order Stationary Points and Excess Risks

DeepPolar: Inventing Nonlinear Large-Kernel Polar Codes via Deep Learning

PLeaS - Merging Models with Permutations and Least Squares

Scalable Fingerprinting of Large Language Models

Improved Communication-Privacy Trade-offs in $L_2$ Mean Estimation under Streaming Differential Privacy

Zonotope Domains for Lagrangian Neural Network Verification

Understanding the Gain from Data Filtering in Multimodal Contrastive Learning

Label Robust and Differentially Private Linear Regression: Computational and Statistical Efficiency

Statistically and Computationally Efficient Linear Meta-representation Learning