Foundation models, i.e. large neural networks pre-trained on large text
...
Pretrained large language models (LLMs) are able to solve a wide variety...
The problem of efficient approximation of a linear operator induced by t...
Knowledge distillation is one of the primary methods of transferring
kno...
We introduce chefs' random tables (CRTs), a new class of non-trigonometr...
Neural language models (LMs) have been shown to memorize a great deal of...
A popular explainable AI (XAI) approach to quantify feature importance o...
The ability to identify influential training examples enables us to debu...
Encoder-decoder transformer architectures have become popular recently w...
There is a set of data augmentation techniques that ablate parts of the ...
We introduce a method called TrackIn that computes the influence of a
tr...
Feature attribution methods, proposed recently, help users interpret the...
In this work, we propose a novel framework for privacy-preserving
client...
Homographs, words with different meanings but the same surface form, hav...
Previous work has modeled the compositionality of words by creating
char...