2 citations · 2 across the 2 of their papers we have counts for
4 papers
Layer-Wise Interpretation of Deep Neural Networks Using Identity Initialization
Shohei Kubota, Hideaki Hayashi, Tomohiro Hayase +1
The interpretability of neural networks (NNs) is a challenging but essential topic for transparency in the decision-making process using machine learning. One of the reasons for th…
Selective Forgetting of Deep Networks at a Finer Level than Samples
Tomohiro Hayase, Suguru Yasutomi, Takashi Katoh
Selective forgetting or removing information from deep neural networks (DNNs) is essential for continual learning and is challenging in controlling the DNNs. Such forgetting is cru…
The Spectrum of Fisher Information of Deep Networks Achieving Dynamical Isometry
Tomohiro Hayase, Ryo Karakida
The Fisher information matrix (FIM) is fundamental to understanding the trainability of deep neural nets (DNN), since it describes the parameter space's local metric. We investigat…
Almost Sure Asymptotic Freeness of Neural Network Jacobian with Orthogonal Weights
Tomohiro Hayase
A well-conditioned Jacobian spectrum has a vital role in preventing exploding or vanishing gradients and speeding up learning of deep neural networks. Free probability theory helps…