Atul Prakash

Atul Prakash

University of Michigan

H-index: 51

North America-United States

About Atul Prakash

Atul Prakash, With an exceptional h-index of 51 and a recent h-index of 23 (since 2020), a distinguished researcher at University of Michigan, specializes in the field of computer security, software systems.

His recent articles reflect a diverse array of research interests and contributions to the field:

PRP: Propagating Universal Perturbations to Attack Large Language Model Guard-Rails

Adaptive Skeleton Graph Decoding

Learn To be Efficient: Build Structured Sparsity in Large Language Models

D4: Detection of Adversarial Diffusion Deepfakes Using Disjoint Ensembles

Stateful Defenses for Machine Learning Models Are Not Yet Secure Against Black-box Attacks

Leveraging Hierarchical Feature Sharing for Efficient Dataset Condensation

Theoretically Principled Trade-off for Stateful Defenses against Query-Based Black-Box Attacks

Concept-based explanations for out-of-distribution detectors

Atul Prakash Information

University

Position

Professor of EECS

Citations(all)

12298

Citations(since 2020)

5807

Cited By

8604

hIndex(all)

51

hIndex(since 2020)

23

i10Index(all)

117

i10Index(since 2020)

44

Email

University Profile Page

Google Scholar

Atul Prakash Skills & Research Interests

computer security

software systems

Top articles of Atul Prakash

PRP: Propagating Universal Perturbations to Attack Large Language Model Guard-Rails

arXiv preprint arXiv:2402.15911

2024/2/24

Adaptive Skeleton Graph Decoding

arXiv preprint arXiv:2402.12280

2024/2/19

Learn To be Efficient: Build Structured Sparsity in Large Language Models

arXiv preprint arXiv:2402.06126

2024/2/9

D4: Detection of Adversarial Diffusion Deepfakes Using Disjoint Ensembles

2024

Stateful Defenses for Machine Learning Models Are Not Yet Secure Against Black-box Attacks

2023/11/15

Leveraging Hierarchical Feature Sharing for Efficient Dataset Condensation

arXiv preprint arXiv:2310.07506

2023/10/11

Theoretically Principled Trade-off for Stateful Defenses against Query-Based Black-Box Attacks

arXiv preprint arXiv:2307.16331

2023/7/30

Concept-based explanations for out-of-distribution detectors

2023

CALICO: Self-Supervised Camera-LiDAR Contrastive Pre-training for BEV Perception

arXiv preprint arXiv:2306.00349

2023/6/1

Investigating Stateful Defenses Against Black-Box Adversarial Examples

arXiv preprint arXiv:2303.06280

2023/3/11

Coverage-centric coreset selection for high pruning rates

arXiv preprint arXiv:2210.15809

2023/1/31

Graphite: Generating automatic physical examples for machine-learning attacks on computer vision systems

2022/6/6

Constraining the Attack Space of Machine Learning Models with Distribution Clamping Preprocessing

arXiv preprint arXiv:2205.08989

2022/5/18

Towards Adversarially Robust Deepfake Detection: An Ensemble Approach

arXiv preprint arXiv:2202.05687

2022/2/11

A Large-scale Investigation into Geodifferences in Mobile Apps

2022

Protecting DNNs from Theft using an Ensemble of Diverse Models

2021/3/6

Sanjay Kariyappa
Sanjay Kariyappa

H-Index: 4

Atul Prakash
Atul Prakash

H-Index: 23

Maze: Data-free model stealing attack using zeroth-order gradient estimation

2021

Sanjay Kariyappa
Sanjay Kariyappa

H-Index: 4

Atul Prakash
Atul Prakash

H-Index: 23

Dispelling Misconceptions and Characterizing the Failings of Deepfake Detection

IEEE Security & Privacy

2021/10/15

Neal Mangaokar
Neal Mangaokar

H-Index: 1

Atul Prakash
Atul Prakash

H-Index: 23

Distributed simulation platforms and data passing tools for natural hazards engineering: Reviews, limitations, and recommendations

2021/10

Simple Run-Time Infrastructure (SRTI): An accessible distributed computing platform for interdisciplinary simulation

Journal of Computational Science

2021/10/1

See List of Professors in Atul Prakash University(University of Michigan)

Co-Authors

academic-engine