Related papers: Distill-and-Compare: Auditing Black-Box Models Usi…
Model distillation is frequently proposed as a technique to reduce the privacy leakage of machine learning. These empirical privacy defenses rely on the intuition that distilled ``student'' models protect the privacy of training data, as…
Algorithmic efficiency techniques such as distillation (\cite{hinton2015distillation}) are useful in improving model quality without increasing serving costs, provided a larger teacher model is available for a smaller student model to learn…
Knowledge distillation leverages a teacher model to improve the training of a student model. A persistent challenge is that a better teacher does not always yield a better student, to which a common mitigation is to use additional…
Recent years have witnessed dramatically improvements in the knowledge distillation, which can generate a compact student model for better efficiency while retaining the model effectiveness of the teacher model. Previous studies find that:…
Ensemble knowledge distillation can extract knowledge from multiple teacher models and encode it into a single student model. Many existing methods learn and distill the student model on labeled data only. However, the teacher models are…
Data-trained predictive models see widespread use, but for the most part they are used as black boxes which output a prediction or score. It is therefore hard to acquire a deeper understanding of model behavior, and in particular how…
Model distillation -- using outputs from a large teacher model to teach a small student model -- is a practical means of creating efficient models for a particular task. We ask: Can we identify a students' teacher based on its outputs? Such…
Knowledge distillation involves transferring the predictive capabilities of large, high-performing AI models (teachers) to smaller models (students) that can operate in environments with limited computing power. In this paper, we address…
The vulnerability of artificial neural networks to adversarial perturbations in the black-box setting is widely studied in the literature. The majority of attack methods to construct these perturbations suffer from an impractically large…
Representation learning has been evolving from traditional supervised training to Contrastive Learning (CL) and Masked Image Modeling (MIM). Previous works have demonstrated their pros and cons in specific scenarios, i.e., CL and supervised…
Active learning can be defined as iterations of data labeling, model training, and data acquisition, until sufficient labels are acquired. A traditional view of data acquisition is that, through iterations, knowledge from human labels and…
We propose a method to infer a dense depth map from a single image, its calibration, and the associated sparse point cloud. In order to leverage existing models (teachers) that produce putative depth maps, we propose an adaptive knowledge…
Compared to large speech foundation models, small distilled models exhibit degraded noise robustness. The student's robustness can be improved by introducing noise at the inputs during pre-training. Despite this, using the standard…
Knowledge distillation is typically conducted by training a small model (the student) to mimic a large and cumbersome model (the teacher). The idea is to compress the knowledge from the teacher by using its output probabilities as…
Knowledge distillation from large language models (LLMs) assumes that the teacher's output distribution is a high-quality training signal. On reasoning tasks, this assumption is frequently violated. A model's intermediate representations…
In the past year, distillation has seen a renewed prominence in large language model (LLM) pretraining, exemplified by the Llama-3.2 and Gemma model families. While distillation has historically been shown to improve statistical modeling,…
Knowledge distillation is a widely used technique for model compression. We posit that the teacher model used in a distillation setup, captures relationships between classes, that extend beyond the original dataset. We empirically show that…
Knowledge distillation (KD) is an efficient approach to transfer the knowledge from a large "teacher" network to a smaller "student" network. Traditional KD methods require lots of labeled training samples and a white-box teacher…
Generalization Performance of Deep Learning models trained using Empirical Risk Minimization can be improved significantly by using Data Augmentation strategies such as simple transformations, or using Mixed Samples. We attempt to…
Knowledge distillation in neural networks refers to compressing a large model or dataset into a smaller version of itself. We introduce Privacy Distillation, a framework that allows a text-to-image generative model to teach another model…