When LLMs directly drive NPC dialogue in social deduction games, a single information leak collapses the entire puzzle structure, a consequence far more severe than in casual social simulations. This paper presents a three-layer hybrid architecture that relocates LLM involvement to a narrow intent classification boundary, while delegating social state evolution to the Ensemble social physics engine and natural language output to a RiveScript dialogue module. A turn-taking mechanism integrating adjacency pair rules and probabilistic volition-weighted bidding coordinates multi-NPC discussion without central LLM mediation. Evaluation through dialogue breakdown analysis across 20 sessions (598 turns) reveals that repetition accounts for 60% of breakdowns, traceable to a structural bottleneck where the scripting layer collapses the engine’s fine-grained action variants into shallow response pools. An adapted Bi-Fact evaluation of the NLU layer confirms that the narrow-channel design achieves stable factual accuracy at the cost of systematic semantic compression. Playtesting further surfaces an emergent genre shift: players reorient from evidence-based deduction toward social relationship management, suggesting that social physics engines may constitute a generative substrate for a distinct interactive genre.

We propose TensionFab, a novel technique for creating shape-changeable room-scale structures. This method employs easily available planar materials (plywood, MDF), cuts them into multiple 2D shapes, and then connects the pieces manually to create a unified structure capable of 2D and 3D deformations. Constructed TensionFab structures are characterized by easy achievability of target surfaces, substantial structural strength, shape changeability, time and material savings, and easy storage and transportation. In this paper, we introduce the basic principles of shape-making, module combination, and structural characteristics of TensionFab. We also developed a design assistance tool to allow users to automate design based on the target shape. We evaluate the design results for shape validation and structural performance. Finally, we validate the approach with actual construction and propose a variety of application scenarios. Overall, TensionFab is an efficient strategy for spatial design and structural organization. This paper contributes to research and exploration in the HCI room-scale interaction field.

Whispers from Silken Margins delves into the transformative silk production techniques of the Showa era, marking a shift from the traditional volumetric cocoon artistry to an avant-garde planar structure. Beyond its inherent promise of sustainability and preservation of both silkworm life and natural resources, this approach heralds a novel dimension in silk artistry. Meticulous studies on silk fiber structures, combined with their inherent stability and plasticity, intertwine with modern architectural methodologies. The resultant aesthetic carves a distinctive niche in three-dimensional space, culminating in an installation that draws inspiration from Archimedean solids. Such a design becomes emblematic of the silk’s metamorphic journey, encapsulating life’s vigor and the overarching creative spirit.

Inflatable systems have been attracting attention in the field of interaction design. Conventional tabletop-sized pneumatic systems tend to be complex because they require bulky and noisy equipment. Therefore, several liquid-to-gas phase change actuators that use vaporization have been proposed. But these actuators have problems with controllability, reusability, and reconfigurability. In this study, we propose InflatableMod, novel inflatable modules based on the efficient control of liquid-to-gas phase change actuators. These are designed with a compact circuit that has a liquid transfer function to feed the required amount of low-boiling-point liquid into the pouch and a heating function to inflate the pouch by the volume change. This approach allows for a compact, silent, and untethered inflatable system. It is also possible to create an untethered and reconfigurable multi-inflatable system because each module is synchronized. In this paper, we propose the design of the modules, evaluate their performance, and present application scenarios.

Physical interfaces that change appearance by controlling transparency have attracted attention for wide applicability. However, many existing methods are limited to planar objects, and it is difficult to control transparency in three-dimensional objects and to control transparency continuously. In this study, we propose a novel method to control the transparency of three-dimensional objects by utilizing the mechanical emulsification phenomenon of liquid. This approach is to fill the object with two transparent liquids that are liquid-liquid phase separated and agitate the liquids inside the object to make them cloudy and control their transparency. By using electrohydrodynamic (EHD) pumps, which can be freely positioned inside the object, to agitate the liquid, it is possible to control the transparency of complex three-dimensional objects. Continuous control is also achieved by changing the ratio of the two liquids. In this paper, we describe the details of proposed method for controlling the transparency and show application scenarios.