Nishikawa, Naoki

8 publications

NeurIPS 2025 Degrees of Freedom for Linear Attention: Distilling SoftMax Attention with Optimal Feature Efficiency Naoki Nishikawa, Rei Higuchi, Taiji Suzuki
NeurIPS 2025 From Shortcut to Induction Head: How Data Diversity Shapes Algorithm Selection in Transformers Ryotaro Kawata, Yujin Song, Alberto Bietti, Naoki Nishikawa, Taiji Suzuki, Samuel Vaiter, Denny Wu
ICML 2025 Mixture of Experts Provably Detect and Learn the Latent Cluster Structure in Gradient-Based Learning Ryotaro Kawata, Kohsei Matsutani, Yuri Kinoshita, Naoki Nishikawa, Taiji Suzuki
ICML 2025 Nonlinear Transformers Can Perform Inference-Time Feature Learning Naoki Nishikawa, Yujin Song, Kazusato Oko, Denny Wu, Taiji Suzuki
ICLR 2025 State Space Models Are Provably Comparable to Transformers in Dynamic Token Selection Naoki Nishikawa, Taiji Suzuki
ICMLW 2024 State Space Models Are Comparable to Transformers in Estimating Functions with Dynamic Smoothness Naoki Nishikawa, Taiji Suzuki
NeurIPS 2023 Adaptive Topological Feature via Persistent Homology: Filtration Learning for Point Clouds Naoki Nishikawa, Yuichi Ike, Kenji Yamanishi
NeurIPS 2022 Two-Layer Neural Network on Infinite Dimensional Data: Global Optimization Guarantee in the Mean-Field Regime Naoki Nishikawa, Taiji Suzuki, Atsushi Nitanda, Denny Wu