activity
20242026
collaborators

5 papers

cs.LG2026

Simulation-free and finite-time diffusion model

Kentaro Kaba, Masayuki Ohzeki, Yuki Sughiyama

The performance of generative diffusion models is determined by the choice of the reference diffusion process connecting the empirical and prior distributions. Conventional approac…

q-bio.MN2025

Information geometry of chemical reaction networks: Cramer-Rao bound and absolute sensitivity revisited

Dimitri Loutchko, Yuki Sughiyama, Tetsuya J. Kobayashi

Information geometry is based on classical Legendre duality but allows to incorporate additional structure such as algebraic constraints and Bregman divergence functions. It is nat…

q-bio.MN2025

Cramer-Rao bound and absolute sensitivity in chemical reaction networks

Dimitri Loutchko, Yuki Sughiyama, Tetsuya J. Kobayashi

Chemical reaction networks (CRN) comprise an important class of models to understand biological functions such as cellular information processing, the robustness and control of met…

cs.LG2024

Schödinger Bridge Type Diffusion Models as an Extension of Variational Autoencoders

Kentaro Kaba, Reo Shimizu, Masayuki Ohzeki +1

Generative diffusion models use time-forward and backward stochastic differential equations to connect the data and prior distributions. While conventional diffusion models (e.g.,…

cond-mat.stat-mech2024

Thermodynamic and Stoichiometric Laws Ruling the Fates of Growing Systems

Atsushi Kamimura, Yuki Sughiyama, Tetsuya J. Kobayashi

We delve into growing open chemical reaction systems (CRSs) characterized by autocatalytic reactions within a variable volume, which changes in response to these reactions. Underst…