Mining Knowledge from Texts: AI Models for Heritage Preservation
A new workflow automatically extracts material properties from scientific literature, offering a powerful tool for more accurate and efficient cultural heritage conservation.
Science
Living Machines: Predicting Biohybrid Actuator Performance with Machine Learning
![The system employed four distinct waveform patterns to deliver electrical stimulation via a portable device, inducing contraction within a biohybrid muscle actuator positioned between movable pillars-a configuration designed to translate modulated [latex] ramp [/latex] or [latex] staircase [/latex] patterns of stimulation into observable mechanical movement.](https://arxiv.org/html/2602.16330v1/fig_ric.png)
Researchers are leveraging the power of machine learning to accurately forecast the behavior of actuators built from living muscle tissue and flexible materials.
Beyond Accuracy: The Reliability Gap in AI Agents
![Despite gains in model utility, reliability metrics demonstrate limited overall improvement, with predictability and safety on [latex]\tau\tau[/latex]-bench, and robustness on GAIA, standing as notable exceptions to this trend.](https://arxiv.org/html/2602.16666v1/x9.png)
As artificial intelligence systems become increasingly capable, a critical challenge is emerging: ensuring their behavior is consistent, robust, and predictable.
Beyond Automation: Can AI Truly Partner with Humans?
A new review examines the evolving relationship between people and artificial intelligence, exploring how to build collaborative decision-making systems that move beyond simple assistance.
Bridging the Gap: AI Agents Now Communicate Directly with Humans
A new protocol establishes a standardized communication framework that integrates people as active nodes within AI agent networks, fostering seamless collaboration and resolving ambiguities.
RoboGene: Teaching Robots to Learn by Doing
![RoboGene cultivates a system where task generation emerges from the interplay of exploratory sampling-guided by a Least Frequently Used [latex] LFU [/latex] strategy-and iterative refinement through self-reflection, all while anchoring performance to sustained learning from Human-in-the-Loop [latex] HITL [/latex] feedback within a long-term memory module.](https://arxiv.org/html/2602.16444v1/x1.png)
A new framework automatically generates realistic tasks for robots, dramatically improving their ability to understand and interact with the physical world.
Beyond the Black Box: Empowering Teachers to Tackle AI Justice
New research reveals how equipping high school computer science educators with practical AI auditing tools fosters critical thinking and a sense of agency in addressing algorithmic harms.
Seeing and Grasping: Humanoid Robots Learn to Manipulate the World
Researchers have developed a new system enabling humanoid robots to reliably pick up and move novel objects in unfamiliar environments using only visual input.
Beyond Black Boxes: A New Framework for Understanding Abstraction
A novel category-theoretic approach unifies different forms of abstraction, offering a rigorous way to build and interpret complex systems.
Reading Minds, Building Robots: Understanding Human Intent
A new review synthesizes psychological insights into human intention, offering a framework to advance collaborative robotics and improve how machines understand our goals.