This article presents ENACT, a framework for assessing embodied cognition using egocentric interaction world modeling. It discusses key findings from various modeling tasks, highlighting performance gaps between models and human capabilities, as well as biases in visual processing. The research emphasizes the limitations of current models in mobile manipulation contexts.

MolmoAct is an innovative Action Reasoning Model (ARM) developed to enhance spatial reasoning in robotics, allowing machines to understand and execute tasks in three-dimensional space. Built on the open-source Molmo framework, MolmoAct utilizes depth-aware perception tokens for improved action planning and execution, demonstrating superior performance and generalization capabilities in real-world scenarios. The model is fully open-source, promoting transparency and accessibility for further research and development in the field.

General Intuition has secured $134 million in seed funding to develop artificial intelligence agents capable of improving spatial reasoning through the analysis of video game clips. This innovative approach aims to enhance the learning capabilities of AI by leveraging the dynamic environments found in gaming. The funding will support the company’s efforts in refining these agents for various applications, including robotics and autonomous systems.