Defining and quantifying the emergence of sparse concepts in dnns
This paper aims to illustrate the concept-emerging phenomenon in a trained DNN.
Specifically, we find that the inference score of a DNN can be disentangled into the effects of …
Specifically, we find that the inference score of a DNN can be disentangled into the effects of …
Loss balancing for fair supervised learning
MM Khalili, X Zhang… - … Conference on Machine …, 2023 - proceedings.mlr.press
Supervised learning models have been used in various domains such as lending, college
admission, face recognition, natural language processing, etc. However, they may inherit …
admission, face recognition, natural language processing, etc. However, they may inherit …
[PDF][PDF] Defining and Quantifying the Emergence of Sparse Concepts in DNNs: Supplementary Material
J Ren, M Li, Q Chen, H Deng, Q Zhang - openaccess.thecvf.com
Explanations for DNNs. Many methods have been proposed to explain DNNs, such as
visualizing the features learned by the DNN [16, 46, 62, 63], and estimating the pixel-wise …
visualizing the features learned by the DNN [16, 46, 62, 63], and estimating the pixel-wise …