Igniting Curiosity: How to Build AI Tutors That Learn What Students Love
New research explores the art of weaving student interests into personalized learning, offering crucial insights for designing truly engaging AI-powered educational tools.
Science
Robots Learn to Build Without Instructions
A new reinforcement learning framework empowers robots to autonomously assemble stable structures from individual blocks, bypassing the need for pre-programmed plans.
Robots Learn by Watching: A New Path to Skill Acquisition
A new approach allows robots to master complex manipulation tasks simply by observing human demonstrations, bypassing the need for explicit programming or reward signals.
Beyond the Algorithm: Teens and the Future of AI-Powered Health
New research reveals how adolescents envision artificial intelligence supporting their health journeys, prioritizing understanding and control over simple efficiency.
The Surgical Second: Humanoid Robots Step into the Operating Room
A new study explores the potential of teleoperated humanoid robots to provide stable endoscopic visualization and versatile assistance during surgical procedures.
Beyond Black Boxes: Reasoning About Machine Learning Explanations
![The system navigates the inherent limitations of base decision models-whether directly trained ([latex]DT[/latex]), globally approximated, or locally surrogated-by generating explanations constrained not by the model itself, but by a meta-interpretation of query language, informed by both user input characteristics (background and distance) and the model’s own embedded representations, acknowledging that any explanation is fundamentally a prophecy of future inadequacy.](https://arxiv.org/html/2602.23810v1/2602.23810v1/workflow.png)
A new framework allows users to not just see why a model made a decision, but to actively reason about those explanations and explore alternative scenarios.
Shaping Soft Robotics: Actuators Built on Geometry
A new approach to pneumatic actuator design leverages geometric principles and constraint layers to achieve precise and reliable soft robotic movement.
Predictive AI: Building Agents That Learn and Adapt with Confidence
![An agent learns to navigate a complex world not by directly modeling its dynamics, but by constructing a verifiable world model-a learned representation assessed by a dedicated verifier-that simultaneously optimizes performance and guarantees adherence to a user-defined specification [latex]\varphi[/latex], effectively decoupling policy learning from precise environmental knowledge and enabling runtime certification of both behavioral correctness and model abstraction quality.](https://arxiv.org/html/2602.23997v1/2602.23997v1/x1.png)
A new framework combines reinforcement learning with formal verification to create AI agents capable of reliable performance in dynamic, real-world environments.
Better Moves: Optimizing Robotic Control Through Action Space Design
A comprehensive study reveals how the way robots are told to move significantly impacts their performance and ability to generalize to new tasks.
AI in the Courtroom: Navigating the Future of Legal Disputes
A new analysis explores how generative AI technologies could reshape legal conflict resolution, presenting both opportunities and challenges for the justice system.