MazeMastery is a Python framework for teaching maze traversal to high school students. It helps students develop abstraction skills by providing a didactic tool. They can test their algorithms against randomized test cases of increasing complexity. The framework facilitates learning analysis and conceptual challenges. MazeMastery is an open-source project for scientists and educators.

RLHF has the potential to transform the way we interact with RL agents. However, the current state of the art in RLHF is limited by the lack of a unified and configurable interface for human feedback. To address this, we present RLHF-Blender, a configurable interactive interface for learning from diverse human feedback. RLHF-Blender is designed to be a flexible and extensible platform that can be easily adapted to different RLHF scenarios.

Integrating empirical evidence into architectural design has the potential to create more humane, legible, and inclusive spaces. Yet, the fast-paced nature of the design process, limited proficiency in scientific terminology, and a natural reliance on design intuition limit architects’ ability to effectively apply empirical findings to inform their design decisions. To address this gap, this paper presents DesignMind, a toolkit for evidence-based, cognitively-informed, and human-centred architectural design. DesignMind has four complementary modules: (1) evidence-based design flashcards, (2) topological and geometric analysis, (3) agent-based simulation, and (4) cognitive walkthrough in virtual reality.