Transformer networks have revolutionized NLP representation learning since they were introduced. Though a great effort has been made to explain the representation in …
In high-throughput functional genomic screens, each gene product is commonly assumed to exhibit a singular biological function within a defined protein complex or pathway. In …
R Huben, H Cunningham, LR Smith… - The Twelfth …, 2023 - openreview.net
One of the roadblocks to a better understanding of neural networks' internals is\textit {polysemanticity}, where neurons appear to activate in multiple, semantically distinct …
We describe a minimalistic and interpretable method for unsupervised learning, without resorting to data augmentation, hyperparameter tuning, or other engineering designs, that …
Flow-based generative models have become an important class of unsupervised learning approaches. In this work, we incorporate the key ideas of renormalization group (RG) and …
We describe a minimalistic and interpretable method for unsupervised representation learning that does not require data augmentation, hyperparameter tuning, or other …
Q Wang, MJ Zaki - … of the 29th International Conference on …, 2022 - aclanthology.org
Learning word embeddings is an essential topic in natural language processing. Most existing works use a vast corpus as a primary source while training, but this requires …
A Templeton - arXiv preprint arXiv:2004.13847, 2020 - arxiv.org
Word embeddings are a powerful natural language processing technique, but they are extremely difficult to interpret. To enable interpretable NLP models, we create vectors where …
W PEIXU, C Yu, Y Ming - ICML 2024 Workshop on Mechanistic … - openreview.net
Large pretrained language models have demonstrated impressive capabilities, but there is still much to learn about how they operate mechanically. In this study, we conduct a …