Critical learning periods are periods early in development where tempora...
We introduce Compartmentalized Diffusion Models (CDM), a method to train...
We describe a first step towards learning general-purpose visual
represe...
We investigate whether prompts learned independently for different tasks...
We present Synergy Aware Forgetting Ensemble (SAFE), a method to adapt l...
Recovering the latent factors of variation of high dimensional data has ...
Responsible use of data is an indispensable part of any machine learning...
We introduce Train/Test-Time Adaptation with Retrieval (T^3AR), a
method...
We propose InCA, a lightweight method for transfer learning that
cross-a...
We propose a novel deterministic method for preparing arbitrary quantum
...
We propose an approach to estimate the number of samples required for a ...
We investigate compositional structures in vector data embeddings from
p...
We introduce À-la-carte Prompt Tuning (APT), a transformer-based scheme ...
We propose a continual learning method which incorporates information fr...
We show that the ability of a neural network to integrate information fr...
We revisit the classic signal-to-symbol barrier in light of the remarkab...
We derive information theoretic generalization bounds for supervised lea...
We propose a notion of common information that allows one to quantify an...
Adapting pre-trained models with broad capabilities has become standard
...
Memorization of the relation between entities in a dataset can lead to
p...
We introduce AdaMix, an adaptive differentially private algorithm for
tr...
We present a method to compute the derivative of a learning task with re...
Fine-tuning from a collection of models pre-trained on different domains...
We define a notion of information that an individual sample provides to ...
We propose a new framework, Translation between Augmented Natural Langua...
We show that the influence of a subset of the training samples can be re...
Classifiers that are linear in their parameters, and trained by optimizi...
We introduce a notion of usable information contained in the representat...
We tackle the problem of predicting the number of optimization steps tha...
Recent results show that features of adversarially trained networks for
...
We address the problem of layout generation for diverse domains such as
...
We describe a procedure for removing dependency on a cohort of training ...
We present a detector for curved text in natural images. We model scene ...
We explore the problem of selectively forgetting a particular set of dat...
We explore the problem of selectively forgetting a particular set of dat...
We study the relationship between catastrophic forgetting and properties...
Regularization is typically understood as improving generalization by
al...
Whatever information a Deep Neural Network has gleaned from past data is...
We introduce an asymmetric distance in the space of learning tasks, and ...
We introduce a method to provide vectorial representations of visual
cla...
We study the topology of the space of learning tasks, which is critical ...
Intelligent behaviour in the real-world requires the ability to acquire ...
Critical periods are phases in the early development of humans and anima...
We review the problem of defining and inferring a "state" for a control
...
Using established principles from Information Theory and Statistics, we ...
The cross-entropy loss commonly used in deep learning is closely related...