Ibaraki University

Abstract Gyrotrons having a high output-frequency of more than 200 GHz will be necessary for future fusion devices, such as the DEMO reactor. At 170 GHz, gyrotrons for ITER achieved a high power of 1 MW with a high electrical efficiency of approximately 50% (oscillation efficiency of approximately 30%) in a long pulse operation. At frequencies higher than 200 GHz, no gyrotrons have achieved ITER-like performance, which this study attempted to demonstrate using a prototype multi-frequency gyrotron for 170 GHz, 203 GHz, and 236 GHz. At 203 GHz, an output power of 0.87 MW was achieved at the gyrotron window with an oscillation efficiency of 25% and an electrical efficiency of 41% in a long pulse operation of 100 s. Although this performance is lower than that of the ITER gyrotrons at 170 GHz, this study represents the first demonstration that ITER-like performance is achievable at frequencies above 200 GHz. In addition, short 1-ms pulse tests at 236 GHz achieved an output power of 1 MW.

The Cover Feature illustrates π-conjugated polymers reducing oxygen (O2) to produce hydrogen peroxide (H2O2), driven by light and electrical energy. The naphthalene units introduced in place of benzene units resulted in the selective deepening of the HOMO energy level of the π-conjugated polymer. The degree of HOMO energy level deepening quantitatively affected the onset potential for light-assisted O2 reduction toward more positive values, enhancing the photo-electrocatalytic H2O2 production rate. More information can be found in the Research Article by K. Oka and co-workers (DOI: 10.1002/cssc.202502396).

Complex infrastructure, as well as economic and social systems, are often represented as networks. The continuedoperation of most networks is threatened by random failures and targeted attacks. Therefore, quantifying andimproving network robustness remains a fundamental challenge in network science. The optimal percolation probleminvolves identifying the smallest set of nodes whose removal would most rapidly fragment the network. Recent evidencesuggests that the presence of redundant cycles significantly restricts optimal solutions. Building on this insight,we propose the Cycle-Redundancy Score (CRS), a purely local metric that estimates the number of independent cyclessupported by a node. We also design a CRS-guided attack algorithm. We tested the robustness of networks againstCRS-guided attacks on both synthetic and empirical networks. On synthetic networks, CRS dismantles networks asquickly as CoreHD. However, on empirical networks with pronounced clustering, CRS systematically underperformscompared to CoreHD and Collective-Influence attacks. A comparative analysis reveals that strong local clustering effectivelyprotects important nodes from early removal, thereby limiting the effectiveness of CRS. These results clarifythe circumstances in which cycle-based dismantling is advantageous and emphasize the necessity of attack strategiesthat can adapt to mesoscale clustering.

We investigated the role of position-dependent memorability and input dynamics in password recall tasks. Eight participants performed a 2 (length: 6-character vs. 8-character) × 4 (character set: numeric/alphabetic/alphanumeric/alphanumeric + symbols) task with physiological and behavioral measures (EEG, blink index, keystroke logs). Generalized estimating equations (GEE) confirmed a robust sequential position effect: accuracy declined toward later positions in the 6-character sequence, with this trend being even more pronounced in the 8-character sequence. The immediate interkey interval (IKI) negatively predicted character-level accuracy, suggesting that local reaction delay serves as a proximal marker for recall failure. In contrast, simple Spearman correlations between accuracy, self-reported fatigue, EEG band power, blink indicators, and reverse digit span were small and non-significant. These findings reaffirm the universality of sequential position mechanisms in mixed-character stimuli and identify IKI as a practical behavioral metric for interface design. We discuss creating passwords that balance usability and robustness.

Abstract The Indo‐Pacific Gateway is a narrow tropical passage connecting the Pacific and Indian Oceans, and alterations in its topographic configuration can significantly influence the volume of water exchanged between the two oceans through the Indonesian Throughflow (ITF). In the modern ocean, the ITF extends from the surface to a depth of 500 m, with its strongest flow occurring in the upper thermocline (100–300 m). Here, we present radiolarian assemblage data from Ocean Drilling Program Site 710 in the western tropical Indian Ocean to reconstruct the history of ITF variability since the Late Miocene. The radiolarian assemblages associated with the ITF can be categorized into two groups: (a) surface ITF taxa, linked to shallow waters, and (b) thermocline ITF taxa, which correspond to subsurface waters. A significant decrease in the relative abundance of surface ITF taxa occurred at 7.5 Ma. Additionally, a faunal turnover from high‐salinity thermocline taxa to low‐salinity thermocline ITF taxa was observed at 3.5 Ma, suggesting a paleoceanographic transition from higher to lower salinity conditions in the thermocline. We propose that these changes reflect modifications in the topographic configuration of the Indo‐Pacific Gateway, which likely occurred in a stepwise manner over these periods. Radiolarian‐based SST data indicate stepwise cooling from 9.0 to 6.5 Ma, followed by a reversal to a warming trend from 4.5 to 2.0 Ma, with a slight decrease in SST after 2.0 Ma. These SST variations are likely influenced by regional topographic changes such as ITF dynamics, in addition to global climate trends.

We present a learning-support system that integrates physical three-dimensional models with augmented reality (AR) to facilitate the acquisition of spatial-figure concepts. The system is implemented on an iPad, which functions as both input and output device. Following instructions displayed on the screen, users mentally construct the spatial figure that satisfies a displayed problem and then place vertex markers along the edges of a physical cube model. The iPad's rear camera tracks these markers, and the system superimposes the user-defined spatial figure onto the model in real-time AR, enabling the learner to examine the figure from multiple viewpoints. While interacting with the system, data-including responses, completion times, and pre- and post-task questionnaire scores-are recorded. These data are subsequently analyzed to evaluate learning outcomes and to inform iterative refinement of instructional content and exhibit design. An evaluation experiment was conducted at a university festival. Nineteen participants completed questionnaires assessing ease of use, perceived effectiveness for learning spatial figures, and interest. The results indicated generally positive evaluations. Although the system did not yield immediate, statistically significant learning gains-likely due to operational difficulty and the inherent challenge of the material-the integration of AR with tangible models appears promising for stimulating learner engagement in the comprehension of spatial figures.