research
∙
08/22/2023
Ceci n'est pas une pomme: Adversarial Illusions in Multi-Modal Embeddings
Multi-modal encoders map images, sounds, texts, videos, etc. into a sing...
share
research
∙
07/19/2023
(Ab)using Images and Sounds for Indirect Instruction Injection in Multi-Modal LLMs
We demonstrate how images and sounds can be used for indirect prompt and...
research
∙
02/09/2023
Hyperparameter Search Is All You Need For Training-Agnostic Backdoor Robustness
Commoditization and broad adoption of machine learning (ML) technologies...
research
∙
08/29/2022
Data Isotopes for Data Provenance in DNNs
Today, creators of data-hungry deep neural networks (DNNs) scour the Int...
research
∙
12/09/2021
Spinning Language Models for Propaganda-As-A-Service
We investigate a new threat to neural sequence-to-sequence (seq2seq) mod...
research
∙
11/17/2021
BigFoot: Exploiting and Mitigating Leakage in Encrypted Write-Ahead Logs
Modern databases and data-warehousing systems separate query processing ...
research
∙
07/22/2021
Spinning Sequence-to-Sequence Models with Meta-Backdoors
We investigate a new threat to neural sequence-to-sequence (seq2seq) mod...
research
∙
11/09/2020
Adversarial Semantic Collisions
We study semantic collisions: texts that are semantically unrelated but ...
research
∙
07/05/2020
You Autocomplete Me: Poisoning Vulnerabilities in Neural Code Completion
Code autocompletion is an integral feature of modern code editors and ID...
research
∙
06/17/2020
De-Anonymizing Text by Fingerprinting Language Generation
Components of machine learning systems are not (yet) perceived as securi...
research
∙
05/08/2020
Blind Backdoors in Deep Learning Models
We investigate a new method for injecting backdoors into machine learnin...
research
∙
02/12/2020
Salvaging Federated Learning by Local Adaptation
Federated learning (FL) is a heavily promoted approach for training ML m...
research
∙
01/14/2020
Humpty Dumpty: Controlling Word Meanings via Corpus Poisoning
Word embeddings, i.e., low-dimensional vector representations such as Gl...
research
∙
05/28/2019
Differential Privacy Has Disparate Impact on Model Accuracy
Differential privacy (DP) is a popular mechanism for training machine le...
research
∙
05/28/2019
Overlearning Reveals Sensitive Attributes
`Overlearning' means that a model trained for a seemingly simple objecti...
research
∙
11/01/2018
The Natural Auditor: How To Tell If Someone Used Your Words To Train Their Model
To help enforce data-protection regulations such as GDPR and detect unau...
research
∙
07/02/2018
How To Backdoor Federated Learning
Federated learning enables multiple participants to jointly construct a ...
research
∙
05/10/2018
Inference Attacks Against Collaborative Learning
Collaborative machine learning and related techniques such as distribute...
research
∙
03/15/2018
Chiron: Privacy-preserving Machine Learning as a Service
Major cloud operators offer machine learning (ML) as a service, enabling...
research
∙
02/15/2018
Fooling OCR Systems with Adversarial Text Images
We demonstrate that state-of-the-art optical character recognition (OCR)...
research
∙
10/18/2016
Membership Inference Attacks against Machine Learning Models
We quantitatively investigate how machine learning models leak informati...
research
∙
09/01/2016
Defeating Image Obfuscation with Deep Learning
We demonstrate that modern image recognition methods based on artificial...
research
∙
02/14/2016
