Related papers: Data Provenance via Differential Auditing
Data-poisoning backdoor attacks are serious security threats to machine learning models, where an adversary can manipulate the training dataset to inject backdoors into models. In this paper, we focus on in-training backdoor defense, aiming…
Query-based black-box attacks have emerged as a significant threat to machine learning systems, where adversaries can manipulate the input queries to generate adversarial examples that can cause misclassification of the model. To counter…
We analyze the data-dependent capacity of neural networks and assess anomalies in inputs from the perspective of networks during inference. The notion of data-dependent capacity allows for analyzing the knowledge base of a model populated…
Constructing a differentially private (DP) estimator requires deriving the maximum influence of an observation, which can be difficult in the absence of exogenous bounds on the input data or the estimator, especially in high dimensional…
Public research results on large-scale supervised finetuning of AI agents remain relatively rare, since the collection of agent training data presents unique challenges. In this work, we argue that the bottleneck is not a lack of underlying…
Distribution shift between train (source) and test (target) datasets is a common problem encountered in machine learning applications. One approach to resolve this issue is to use the Unsupervised Domain Adaptation (UDA) technique that…
We present a method to compute the derivative of a learning task with respect to a dataset. A learning task is a function from a training set to the validation error, which can be represented by a trained deep neural network (DNN). The…
The impact of inference-time data perturbation (e.g., adversarial attacks) has been extensively studied in machine learning, leading to well-established certification techniques for adversarial robustness. In contrast, certifying models…
Differential Privacy (DP) is widely adopted in data management systems to enable data sharing with formal disclosure guarantees. A central systems challenge is understanding how DP noise translates into effective protection against…
Malicious adversaries can attack machine learning models to infer sensitive information or damage the system by launching a series of evasion attacks. Although various work addresses privacy and security concerns, they focus on individual…
Regulatory audits of AI systems increasingly rely on differential privacy (DP) to protect training data and model internals. We study audit design when the audited developer can strategically respond to the privacy-constrained audit…
Recent advances in AI-powered image editing tools have significantly lowered the barrier to image modification, raising pressing security concerns those related to spreading misinformation and disinformation on social platforms. Image…
Advanced persistent threats (APT) are stealthy cyber-attacks that are aimed at stealing valuable information from target organizations and tend to extend in time. Blocking all APTs is impossible, security experts caution, hence the…
Fraud detection is to identify, monitor, and prevent potentially fraudulent activities from complex data. The recent development and success in AI, especially machine learning, provides a new data-driven way to deal with fraud. From a…
To protect privacy for data-collection-based services, local differential privacy (LDP) is widely adopted due to its rigorous theoretical bound on privacy loss. However, mistakes in complex theoretical analysis or subtle implementation…
Recent works in self-supervised learning have shown impressive results on single-object images, but they struggle to perform well on complex multi-object images as evidenced by their poor visual grounding. To demonstrate this concretely, we…
Semi-supervised domain adaptation (SSDA) adapts a learner to a new domain by effectively utilizing source domain data and a few labeled target samples. It is a practical yet under-investigated research topic. In this paper, we analyze the…
Diffusion Models (DMs) have emerged as powerful generative models with unprecedented image generation capability. These models are widely used for data augmentation and creative applications. However, DMs reflect the biases present in the…
Reasoning distillation has emerged as a prevailing paradigm for transferring reasoning capabilities from large reasoning models to small language models. Yet, reasoning distillation risks data contamination: benchmark data may inadvertently…
Federated learning (FL) aims to collaboratively train the global model in a distributed manner by sharing the model parameters from local clients to a central server, thereby potentially protecting users' private information. Nevertheless,…