9 citations · 38 across the 15 of their papers we have counts for
15 papers
Removing Spurious Correlation from Neural Network Interpretations
Milad Fotouhi, Mohammad Taha Bahadori, Oluwaseyi Feyisetan +2
The existing algorithms for identification of neurons responsible for undesired and harmful behaviors do not consider the effects of confounders such as topic of the conversation.…
Fast Training Dataset Attribution via In-Context Learning
Milad Fotouhi, Mohammad Taha Bahadori, Oluwaseyi Feyisetan +2
We investigate the use of in-context learning and prompt engineering to estimate the contributions of training data in the outputs of instruction-tuned large language models (LLMs)…
TEM: High Utility Metric Differential Privacy on Text
Ricardo Silva Carvalho, Theodore Vasiloudis, Oluwaseyi Feyisetan
Ensuring the privacy of users whose data are used to train Natural Language Processing (NLP) models is necessary to build and maintain customer trust. Differential Privacy (DP) has…
BRR: Preserving Privacy of Text Data Efficiently on Device
Ricardo Silva Carvalho, Theodore Vasiloudis, Oluwaseyi Feyisetan
With the use of personal devices connected to the Internet for tasks such as searches and shopping becoming ubiquitous, ensuring the privacy of the users of such services has becom…
Reconstructing Test Labels from Noisy Loss Functions
Abhinav Aggarwal, Shiva Prasad Kasiviswanathan, Zekun Xu +2
Machine learning classifiers rely on loss functions for performance evaluation, often on a private (hidden) dataset. In a recent line of research, label inference was introduced as…
Label Inference Attacks from Log-loss Scores
Abhinav Aggarwal, Shiva Prasad Kasiviswanathan, Zekun Xu +2
Log-loss (also known as cross-entropy loss) metric is ubiquitously used across machine learning applications to assess the performance of classification algorithms. In this paper,…
On a Utilitarian Approach to Privacy Preserving Text Generation
Zekun Xu, Abhinav Aggarwal, Oluwaseyi Feyisetan +1
Differentially-private mechanisms for text generation typically add carefully calibrated noise to input words and use the nearest neighbor to the noised input as the output word. W…
Research Challenges in Designing Differentially Private Text Generation Mechanisms
Oluwaseyi Feyisetan, Abhinav Aggarwal, Zekun Xu +1
Accurately learning from user data while ensuring quantifiable privacy guarantees provides an opportunity to build better Machine Learning (ML) models while maintaining user trust.…
A Differentially Private Text Perturbation Method Using a Regularized Mahalanobis Metric
Zekun Xu, Abhinav Aggarwal, Oluwaseyi Feyisetan +1
Balancing the privacy-utility tradeoff is a crucial requirement of many practical machine learning systems that deal with sensitive customer data. A popular approach for privacy-pr…
Differentially Private Adversarial Robustness Through Randomized Perturbations
Nan Xu, Oluwaseyi Feyisetan, Abhinav Aggarwal +2
Deep Neural Networks, despite their great success in diverse domains, are provably sensitive to small perturbations on correctly classified examples and lead to erroneous predictio…
On Primes, Log-Loss Scores and (No) Privacy
Abhinav Aggarwal, Zekun Xu, Oluwaseyi Feyisetan +1
Membership Inference Attacks exploit the vulnerabilities of exposing models trained on customer data to queries by an adversary. In a recently proposed implementation of an auditin…
Leveraging Hierarchical Representations for Preserving Privacy and Utility in Text
Oluwaseyi Feyisetan, Tom Diethe, Thomas Drake
Guaranteeing a certain level of user privacy in an arbitrary piece of text is a challenging issue. However, with this challenge comes the potential of unlocking access to vast data…
Privacy- and Utility-Preserving Textual Analysis via Calibrated Multivariate Perturbations
Oluwaseyi Feyisetan, Borja Balle, Thomas Drake +1
Accurately learning from user data while providing quantifiable privacy guarantees provides an opportunity to build better ML models while maintaining user trust. This paper presen…
Privacy-preserving Active Learning on Sensitive Data for User Intent Classification
Oluwaseyi Feyisetan, Thomas Drake, Borja Balle +1
Active learning holds promise of significantly reducing data annotation costs while maintaining reasonable model performance. However, it requires sending data to annotators for la…
Beyond monetary incentives: experiments in paid microtask contests modelled as continuous-time markov chains
Oluwaseyi Feyisetan, Elena Simperl
In this paper, we aim to gain a better understanding into how paid microtask crowdsourcing could leverage its appeal and scaling power by using contests to boost crowd performance…