Follow
Yuma Ichikawa
Yuma Ichikawa
University of Tokyo, Fujitsu Limited
Verified email at g.ecc.u-tokyo.ac.jp - Homepage
Title
Cited by
Cited by
Year
Statistical-mechanical study of deep boltzmann machine given weight parameters after training by singular value decomposition
Y Ichikawa, K Hukushima
Journal of the Physical Society of Japan 91 (11), 114001, 2022
52022
Adaptive Flip Graph Algorithm for Matrix Multiplication
Y Arai, Y Ichikawa, K Hukushima
arXiv preprint arXiv:2312.16960, 2023
22023
Dataset Size Dependence of Rate-Distortion Curve and Threshold of Posterior Collapse in Linear VAE
Y Ichikawa, K Hukushima
arXiv preprint arXiv:2309.07663, 2023
12023
Learning Dynamics in Linear VAE: Posterior Collapse Threshold, Superfluous Latent Space Pitfalls, and Speedup with KL Annealing
Y Ichikawa, K Hukushima
International Conference on Artificial Intelligence and Statistics, 1936-1944, 2024
2024
Continuous Tensor Relaxation for Finding Diverse Solutions in Combinatorial Optimization Problems
Y Ichikawa, H Iwashita
arXiv preprint arXiv:2402.02190, 2024
2024
Controlling Continuous Relaxation for Combinatorial Optimization
Y Ichikawa
arXiv preprint arXiv:2309.16965, 2023
2023
Toward Unlimited Self-Learning MCMC with Parallel Adaptive Annealing
Y Ichikawa, A Nakagawa, H Masayuki, Y Umeda
arXiv preprint arXiv:2211.14024, 2022
2022
Toward Unlimited Self-Learning Monte Carlo with Annealing Process Using VAE's Implicit Isometricity.
Y Ichikawa, A Nakagawa, H Masayuki, Y Umeda
CoRR, 2022
2022
The system can't perform the operation now. Try again later.
Articles 1–8