Gaze-based selection in XR requires visual confirmation due to eye-tracking limitations and target ambiguity in 3D contexts. Current designs for wide-FOV displays use world-locked, central overlays, which is not conducive to always-on AR glasses. This paper introduces PeriphAR (/peh-ree-faar/), a visualization technique that leverages peripheral vision for feedback during gaze-based selection on a monocular AR display. In a first user study, we isolated text, color, and shape properties of target objects to compare peripheral selection cues. Peripheral vision was more sensitive to color than shape, but this sensitivity rapidly declined at lower contrast. To preserve preattentive processing of color, we developed two strategies to enhance color in users’ peripheral vision. In a second user study, our strategy that maximized contrast of the target to the neighboring object with the most similar color was subjectively preferred. As proof of concept, we implemented PeriphAR in an end-to-end system to test performance with real-world object detection.

Current benchmarks for evaluating the reasoning capabilities of Large Language Models (LLMs) face significant limitations: task oversimplification, data contamination, and flawed evaluation items. These deficiencies necessitate more rigorous assessment methods. To address these limitations, we introduce PHYBench, a benchmark of 500 original physics problems ranging from high school to Physics Olympiad difficulty. PHYBench addresses data contamination through original content and employs a systematic curation pipeline to eliminate flawed items. Evaluations show that PHYBench activates more tokens and provides stronger differentiation between reasoning models compared to other baselines like AIME 2024, OlympiadBench and GPQA. Even the best-performing model, Gemini 2.5 Pro, achieves only 36.9\% accuracy compared to human experts' 61.9\%. To further enhance evaluation precision, we introduce the Expression Edit Distance (EED) Score for mathematical expression assessment, which improves sample efficiency by 204\% over binary scoring. Moreover, PHYBench effectively elicits multi-step and multi-condition reasoning, providing a platform for examining models' reasoning robustness, preferences, and deficiencies. The benchmark results and dataset are publicly available at https://www.phybench.cn/.

Autonomous Guided Vehicles (AGVs) are becoming increasingly common in industrial environments to transport heavy equipment around warehouses. Within the idea of Industry 5.0, these AGVs are expected to work alongside humans in the same shared workspace. To enable smooth and trustworthy interaction between workers and AGVs, the AGVs must be able to model the workers’ behavior and plan their trajectories around it. In this paper, we introduce a Finite Automaton Model (FAM) to model worker motion in such a context. We conduct a human subject experiment using a Virtual Reality (VR) environment and an omnidirectional treadmill to collect data about worker trajectories to tune our model. We show that not only is our model more interpretable, but also outperforms machine learning models at classifying worker motion behavior with limited training data. Future research can use our model to modify AGV behavior to promote trustworthy human-AGV interaction.