Amin Ghiasi — accepted AI-conference papers

2025

Stable Diffusion Models are Secretly Good at Visual In-Context Learning

Trevine Oorloff, Vishwanath Sindagi, Wele Gedara Chaminda Bandara, Ali Shafahi, Amin Ghiasi, Charan Prakash +1

ICCV 2025poster

Large language models (LLM) in natural language processing (NLP) have demonstrated great potential for in-context learning (ICL) -- the ability to leverage a few set of example prompts to adapt to various tasks without having to explicitly update model weights. ICL has recently been explored for the…

Cited by 0 Source ↗PDF ↗Scholar ↗

2023

Improving Robustness with Adaptive Weight Decay

Amin Ghiasi, Ali Shafahi, Reza Ardekani

NeurIPS 2023poster

We propose adaptive weight decay, which automatically tunes the hyper-parameter for weight decay during each training iteration. For classification problems, we propose changing the value of the weight decay hyper-parameter on the fly based on the strength of updates from the classification loss (i.…

Cited by 8 Source ↗PDF ↗Scholar ↗

2022

Plug-In Inversion: Model-Agnostic Inversion for Vision with Data Augmentations

Amin Ghiasi, Hamid Kazemi, Steven Reich, Chen Zhu, Micah Goldblum, Tom Goldstein

ICML 2022spotlight

Existing techniques for model inversion typically rely on hard-to-tune regularizers, such as total variation or feature regularization, which must be individually calibrated for each network in order to produce adequate images. In this work, we introduce Plug-In Inversion, which relies on a simple s…

Cited by 10 Source ↗PDF ↗code ↗Scholar ↗

2022

The Uncanny Similarity of Recurrence and Depth

Avi Schwarzschild, Arjun Gupta, Amin Ghiasi, Micah Goldblum, Tom Goldstein

ICLR 2022poster

It is widely believed that deep neural networks contain layer specialization, wherein networks extract hierarchical features representing edges and patterns in shallow layers and complete objects in deeper layers. Unlike common feed-forward models that have distinct filters at each layer, recurrent…

Cited by 9 Source ↗PDF ↗code ↗Scholar ↗

2021

Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoff

Eitan Borgnia, Valeriia Cherepanova, Liam Fowl, Amin Ghiasi, Jonas Geiping, Micah Goldblum +2

ICASSP 2021accepted

Data poisoning and backdoor attacks manipulate victim models by maliciously modifying training data. In light of this growing threat, a recent survey of industry professionals revealed heightened fear in the private sector regarding data poisoning. Many previous defenses against poisoning either fai…

Cited by 0 Source ↗Scholar ↗

2020

Adversarially robust transfer learning

Ali Shafahi, Parsa Saadatpanah, Chen Zhu, Amin Ghiasi, Christoph Studer, David Jacobs +1

ICLR 2020poster

Transfer learning, in which a network is trained on one task and re-purposed on another, is often used to produce neural network classifiers when data is scarce or full-scale training is too costly. When the goal is to produce a model that is not only accurate but also adversarially robust, data sc…

Cited by 157 Source ↗code ↗Scholar ↗

2020

BREAKING CERTIFIED DEFENSES: SEMANTIC ADVERSARIAL EXAMPLES WITH SPOOFED ROBUSTNESS CERTIFICATES

Amin Ghiasi, Ali Shafahi, Tom Goldstein

ICLR 2020poster

Defenses against adversarial attacks can be classified into certified and non-certified. Certifiable defenses make networks robust within a certain $\ell_p$-bounded radius, so that it is impossible for the adversary to make adversarial examples in the certificate bound. We present an attack that mai…

Cited by 84 Source ↗code ↗Scholar ↗