Folding into Motion: A New Origami-Inspired Motor
Researchers have created the first macro-scale rotary motor built from folded structures and powered by electrostatic forces, opening up new possibilities for deployable robotics.
Science
The Algorithmic Muse: Exploring AI and Sexual Content Creation
A new wave of generative AI tools is enabling unprecedented creative possibilities, but also raising critical ethical questions around consent, safety, and the future of intimate imagery.
Beyond the Algorithm: Building AI-Ready Organizations
Successfully integrating artificial intelligence requires more than just technology – it demands a fundamental shift in organizational practices and a focus on human-centered design.
Designing a Greener Future: AI-Powered Materials Discovery
A new approach combines machine learning with life cycle assessment to accelerate the development of materials that are both high-performing and environmentally sustainable.
Robotic Grasping: Mastering the Unknown in Space
New research demonstrates a pathway to more robust and efficient robotic manipulation in dynamic environments, crucial for off-world operations and beyond.
Decoding Group Theory with Neural Networks
![The study demonstrates that a multilayer perceptron (MLP) trained on the discrete group [latex]D_{30}[/latex] exhibits a structured representation, as evidenced by linear probe accuracy corresponding to alternating and rotational subgroups-a pattern not consistently observed in a transformer network trained on [latex]S_5[/latex], suggesting differing capacities for learning and representing group symmetries within these architectures.](https://arxiv.org/html/2601.21150v1/subgroup_probe_mlp_dihedral.png)
New research explores whether narrow neural networks can learn the underlying algebraic principles of finite groups simply by predicting their operations.
Beyond ROS: Building Scalable Robotics with Meta-ROS
A new middleware architecture aims to unlock the full potential of robotic systems through enhanced communication and cloud integration.
Beyond Prediction: Active Inference Learns to Control Its World
A new framework merges active inference with distributional reinforcement learning, allowing agents to master complex tasks without building explicit world models.
Robots That React: Teaching Machines to Hear and Act
New research explores how robots can move beyond pre-programmed instructions to respond dynamically to real-world events triggered by sound.
Untangling Data: A New Approach to Causal Inference
Researchers have developed a dual-encoding method to stabilize causal discovery when dealing with mixed data types, leading to more reliable and interpretable AI explanations.