Yuma Ichikawa

24 papers A* 2A 1B 1Journal 20
YearRankTypeTitle / Venue / Authors
2026 J jnl
OneComp: One-Line Revolution for Generative AI Model Compression.
CoRR
Yuma Ichikawa, Keiji Kimura, Akihiro Yoshida, Yudai Fujimoto, Hiroki Tokura, Yamato Arai, Yoshiyuki Ishii, Yusei Kawakami, Genki Shikada, Achille Jacquemond, Yoshihiko Fujisawa, Katsuki Fujisawa, Takumi Honda, Akira Sakai
2026 J jnl
Sign Lock-In: Randomly Initialized Weight Signs Persist and Bottleneck Sub-Bit Model Compression.
CoRR
Akira Sakai, Yuma Ichikawa
2026 J jnl
Thermal Min-Max Games: Unifying Bounded Rationality and Typical-Case Equilibrium.
CoRR
Yuma Ichikawa
2025 J jnl
Continuous Parallel Relaxation for Finding Diverse Solutions in Combinatorial Optimization Problems.
Trans. Mach. Learn. Res.
Yuma Ichikawa, Hiroaki Iwashita
2025 J jnl
High-Dimensional Learning Dynamics of Quantized Models with Straight-Through Estimator.
CoRR
Yuma Ichikawa, Shuhei Kashiwamura, Ayaka Sakata
2025 J jnl
LPCD: Unified Framework from Layer-Wise to Submodule Quantization.
CoRR
Yuma Ichikawa, Yudai Fujimoto, Akira Sakai
2025 J jnl
More Than Bits: Multi-Envelope Double Binary Factorization for Extreme Quantization.
CoRR
Yuma Ichikawa, Yoshihiko Fujisawa, Yudai Fujimoto, Akira Sakai, Katsuki Fujisawa
2025 A* conf
Optimization by Parallel Quasi-Quantum Annealing with Gradient-Based Sampling.
ICLR
Yuma Ichikawa, Yamato Arai
2025 J jnl
PHOTON: Hierarchical Autoregressive Modeling for Lightspeed and Memory-Efficient Language Generation.
CoRR
Yuma Ichikawa, Naoya Takagi, Takumi Nakagawa, Yuzi Kanazawa, Akira Sakai
2025 J jnl
Quantization Error Propagation: Revisiting Layer-Wise Post-Training Quantization.
CoRR
Yamato Arai, Yuma Ichikawa
2024 B conf
Adaptive Flip Graph Algorithm for Matrix Multiplication.
ISSAC
Yamato Arai, Yuma Ichikawa, Koji Hukushima
2024 J jnl
Continuous Tensor Relaxation for Finding Diverse Solutions in Combinatorial Optimization Problems.
CoRR
Yuma Ichikawa, Hiroaki Iwashita
2024 A* conf
Controlling Continuous Relaxation for Combinatorial Optimization.
NeurIPS
Yuma Ichikawa
2024 A conf
Learning Dynamics in Linear VAE: Posterior Collapse Threshold, Superfluous Latent Space Pitfalls, and Speedup with KL Annealing.
AISTATS
Yuma Ichikawa, Koji Hukushima
2024 J jnl
Optimization by Parallel Quasi-Quantum Annealing with Gradient-Based Sampling.
CoRR
Yuma Ichikawa, Yamato Arai
2024 J jnl
Ratio Divergence Learning Using Target Energy in Restricted Boltzmann Machines: Beyond Kullback-Leibler Divergence Learning.
CoRR
Yuichi Ishida, Yuma Ichikawa, Aki Dote, Toshiyuki Miyazawa, Koji Hukushima
2024 J jnl
Statistical Mechanics of Min-Max Problems.
CoRR
Yuma Ichikawa, Koji Hukushima
2024 J jnl
Statistical Mechanics of Min-Max Problems.
Trans. Mach. Learn. Res.
Yuma Ichikawa, Koji Hukushima
2024 J jnl
Training-Free Time-Series Anomaly Detection: Leveraging Image Foundation Models.
CoRR
Nobuo Namura, Yuma Ichikawa
2023 J jnl
Adaptive Flip Graph Algorithm for Matrix Multiplication.
CoRR
Yamato Arai, Yuma Ichikawa, Koji Hukushima
2023 J jnl
Controlling Continuous Relaxation for Combinatorial Optimization.
CoRR
Yuma Ichikawa
2023 J jnl
Dataset Size Dependence of Rate-Distortion Curve and Threshold of Posterior Collapse in Linear VAE.
CoRR
Yuma Ichikawa, Koji Hukushima
2023 J jnl
Learning Dynamics in Linear VAE: Posterior Collapse Threshold, Superfluous Latent Space Pitfalls, and Speedup with KL Annealing.
CoRR
Yuma Ichikawa, Koji Hukushima
2022 J jnl
Toward Unlimited Self-Learning Monte Carlo with Annealing Process Using VAE's Implicit Isometricity.
CoRR
Yuma Ichikawa, Akira Nakagawa, Hiromoto Masayuki, Yuhei Umeda