Toru Takahashi

Hiroshi Isakari, Toyohiro Kondo, Toru Takahashi et al.

This paper presents a structural optimisation method in three-dimensional acoustic-elastic coupled problems. The proposed optimisation method finds an optimal allocation of elastic materials which reduces the sound level on some fixed observation points. In the process of the optimisation, configuration of the elastic materials is expressed with a level set function, and the distribution of the level set function is iteratively updated with the help of the topological derivative. The topological derivative is associated with state and adjoint variables which are the solutions of the acoustic-elastic coupled problems. In this paper, the acoustic-elastic coupled problems are solved by a BEM-FEM coupled solver, in which the fast multipole method (FMM) and a multi-frontal solver for sparse matrices are efficiently combined. Along with the detailed formulations for the topological derivative and the BEM-FEM coupled solver, we present some numerical examples of optimal designs of elastic sound scatterer to manipulate sound waves, from which we confirm the effectiveness of the present method.

Toru Takahashi

Mutual misunderstanding in contemporary society does not arise merely because people hold different opinions or values. Even under the same observations, different subjects may form different inferential targets, state representations, prediction errors, and update priorities. This paper proposes a multi-phase inference framework and defines its core internal mechanism as the Multi-Phase Inference Mechanism (MIM). MIM formalizes how heterogeneous world models arise through a phase-formation space, a foregrounding field, subject-specific profile states, and alignment maps between state representations. On this basis, the paper reframes world-model alignment as the problem of making heterogeneous representations mutually processable, rather than forcing agreement or convergence to a single value system. It further connects this formalism to philosophical disagreements, cognitive typology, social fragmentation, and AI alignment. The aim is to provide a constructive vocabulary for AI systems that can help humans understand self and others by making differences in meaning, value, and prediction error visible, comparable, and transformable.

Kenta Nakamoto, Hiroshi Isakari, Toru Takahashi et al.

We show a design method of cloaking devices which work for target objects with arbitrary shape and material by a topology optimisation with an accurate and efficient sensitivity analysis. Most of past researches on topology optimisation of cloaking devices intend to hide a circle-shaped perfect electric conductor. In this case, the cloaking effect is highly dependent on the shape and material of a target object. In this study, we consider to design a cloaking device which work regardless of the property of target objects by modifying the definition of the objective function. Also, we developed an efficient and accurate sensitivity analysis with the boundary element method and the ${\mathcal H}$-matrix method. We show that the proposed method can successfully obtain desired cloaking devices with low computational cost.

Toru Takahashi

When people share the same documents and observations yet reach different conclusions, the disagreement often shifts into a judgment that the other party is cognitively defective, irrational, or acting in bad faith. This paper argues that such divergence is better described as a form of non-identifiability inherent in inference and learning, rather than as a defect of the other party. We organize the phenomenon into two levels: (i) $θ$-level non-identifiability, where conclusions diverge under the same world model $W$ because inference settings differ; and (ii) $W$-level non-identifiability, where repeated use of an inference setting $θ$ biases data exposure and update rules, causing the learned world model $W$ itself to diverge. We introduce an inference profile $θ= (R, E, S, D)$, consisting of Reference, Exploration, Stabilization, and Horizon, and show how outputs can split even for the same observation $o$ and the same $W$. We further explain why disagreements tend to project onto a small number of bases -- abstract versus concrete, externalizability, and order versus freedom -- as a consequence of general constraints on learning systems: computational, observational, and coordination constraints. Finally, we relate the framework to deep representation learning, including representation hierarchy, latent-state estimation, and regularization-exploration trade-offs, and illustrate the framework through a case study on AI regulation debates.