Related papers: On the Robustness Tradeoff in Fine-Tuning

We show that combining human prior knowledge with end-to-end learning can improve the robustness of deep neural networks by introducing a part-based model for object classification. We believe that the richer form of annotation helps guide…

The emerging interest in low-latency high-reliability applications, such as connected vehicles, necessitates a new abstraction between communication and control. Thanks to advances in cyber-physical systems over the past decades, we…

Throughout the past five years, the susceptibility of neural networks to minimal adversarial perturbations has moved from a peculiar phenomenon to a core issue in Deep Learning. Despite much attention, however, progress towards more robust…

As a research community, we are still lacking a systematic understanding of the progress on adversarial robustness which often makes it hard to identify the most promising ideas in training robust models. A key challenge in benchmarking…

The robustness of deep neural networks is usually lacking under adversarial examples, common corruptions, and distribution shifts, which becomes an important research problem in the development of deep learning. Although new deep learning…

Finetuning vision foundation models often improves in-domain accuracy but comes at the cost of robustness under distribution shift. We revisit Mixout, a stochastic regularizer that intermittently replaces finetuned weights with their…

Trade-offs between accuracy and efficiency pervade law, public health, and other non-computing domains, which have developed policies to guide how to balance the two in conditions of uncertainty. While computer science also commonly studies…

Adversarial training has been proven to be an effective technique for improving the adversarial robustness of models. However, there seems to be an inherent trade-off between optimizing the model for accuracy and robustness. To this end, we…

AI systems that model and interact with users can update their models over time to reflect new information and changes in the environment. Although these updates may improve the overall performance of the AI system, they may actually hurt…

Supervised and preference-based fine-tuning techniques have become popular for aligning large language models (LLMs) with user intent and correctness criteria. However, real-world training data often exhibits spurious correlations --…

Adversarial Training (AT) with Projected Gradient Descent (PGD) is an effective approach for improving the robustness of the deep neural networks. However, PGD AT has been shown to suffer from two main limitations: i) high computational…

Adversarial training has been successfully applied to build robust models at a certain cost. While the robustness of a model increases, the standard classification accuracy declines. This phenomenon is suggested to be an inherent trade-off.…

Fine-tuning pre-trained foundation models has gained significant popularity in various research fields. Existing methods for fine-tuning can be roughly divided into two categories, namely Parameter-Efficient Fine-Tuning and High-Performance…

Large models represent a groundbreaking advancement in multiple application fields, enabling remarkable achievements across various tasks. However, their unprecedented scale comes with significant computational costs. These models, often…

Adversarial patches are optimized contiguous pixel blocks in an input image that cause a machine-learning model to misclassify it. However, their optimization is computationally demanding, and requires careful hyperparameter tuning,…

Defenses against adversarial examples, such as adversarial training, are typically tailored to a single perturbation type (e.g., small $\ell_\infty$-noise). For other perturbations, these defenses offer no guarantees and, at times, even…

As multi-task models gain popularity in a wider range of machine learning applications, it is becoming increasingly important for practitioners to understand the fairness implications associated with those models. Most existing fairness…

Recently, large pre-trained foundation models have become widely adopted by machine learning practitioners for a multitude of tasks. Given that such models are publicly available, relying on their use as backbone models for downstream tasks…

In this work, besides improving prediction accuracy, we study whether personalization could bring robustness benefits to backdoor attacks. We conduct the first study of backdoor attacks in the pFL framework, testing 4 widely used backdoor…

Multi-task learning, which optimizes performance across multiple tasks, is inherently a multi-objective optimization problem. Various algorithms are developed to provide discrete trade-off solutions on the Pareto front. Recently, continuous…