Smart Agents, Smarter Budgets: The Future of Conversational AI
New research tackles the challenge of building helpful and affordable virtual assistants that can truly deliver on task-oriented dialogue.
Science
AI Unlocks New Carbon Structures with Tailored Properties
![A generative model, prompted to design stable carbon allotropes, successfully predicted several novel structures-including [latex]C_3\_6[/latex], [latex]C_{24\_4}[/latex], and [latex]C_{52\_{15}}[/latex]-all of which, upon analysis via phonon dispersion relations, exhibited dynamical stability confirmed by the absence of imaginary frequencies, suggesting a predictable link between algorithmic design and material viability.](https://arxiv.org/html/2602.22706v1/2602.22706v1/x5.png)
Researchers have developed an artificial intelligence workflow that designs and predicts the properties of novel carbon allotropes, pushing the boundaries of materials discovery.
Catching on: Robots Learn Agile Manipulation from Pixels Alone
![The deployed policy successfully integrates real-world visual input-specifically, segmented object geometries derived from [latex]SAM2[/latex] analysis of camera feeds-to execute catching sequences, demonstrating an ability to process complex scenes and actuate appropriate responses.](https://arxiv.org/html/2602.22733v1/2602.22733v1/x6.png)
A new framework enables robots to reliably catch dynamic objects using only visual information from a single camera, bypassing the need for complex 3D sensing.
Digital Companions: How AI Influencers Are Shaping Connections for Older Adults
New research explores the growing relationship between seniors and AI-generated personalities on social media, revealing the potential for both emotional support and unforeseen risks.
Bending the Rules: Modeling Precision in Continuum Robotics
![The study models a rod undergoing significant deformation, defining a reference frame [latex]\mathcal{F}_{\tau}[/latex] at each time step τ where the x-axis consistently aligns with the rod’s central axis, enabling precise tracking of its evolving geometry.](https://arxiv.org/html/2602.22854v1/2602.22854v1/x1.png)
New research evaluates the accuracy and efficiency of strain-based models for controlling the complex movements of flexible, snake-like robots.
The Limits of AI Agency
New research reveals fundamental architectural constraints that prevent current AI systems from truly understanding or adhering to human values.
Can AI Conversations Shift the Needle on Climate Action?
![Personalized large language model interventions demonstrated significantly higher impact ranking accuracy and stronger intentions to engage in high-impact climate actions compared to other conditions, as evidenced by statistically significant differences in means and interquartile ranges-effects that remained largely consistent even after applying a conservative Holm correction for multiple comparisons [latex] p < .05 [/latex] and [latex] p < .01 [/latex].](https://arxiv.org/html/2602.22564v1/2602.22564v1/x3.png)
New research suggests personalized interactions with artificial intelligence can correct common misperceptions about effective climate solutions and inspire greater commitment to high-impact behaviors.
Seeing Beyond the Sensor: Robust Object Detection with Event Cameras
New research shows that training event-based vision systems with varied simulated sensor characteristics dramatically improves their adaptability to real-world conditions.
Beyond Apps: Giving Smartphones a Mind of Their Own
New research explores a system that allows smartphones to perform complex tasks autonomously, bridging the gap between language commands and real-world device control.
The Shape of Life’s Data: What Foundation Models Really Learn
A new study systematically maps the geometric landscape within single-cell genomic foundation models, revealing that apparent biological structure often lacks robustness.