Navigating the Long Read: Smarter Document Search with AI Agents
A new framework improves question answering over lengthy documents by understanding their structure and intelligently navigating content.
Science
Understanding What Users *Really* Want: A New Approach to Recommendation
Researchers have developed a system that uses the power of large language models to decipher user motivations from reviews and behavior, leading to more relevant and effective recommendations.
Uncovering Hidden Drivers in Continuous-Time Data
New research bridges causal inference and stochastic modeling to reveal the underlying mechanisms generating complex event sequences.
Can Robots Truly Grasp the Basics?
![Despite all BusyBox configurations being within the training data’s affordance distribution, the algorithms [latex]\pi_{0.5}\pi_{0.5}-canon[/latex] and GR00T-N1.6-canon demonstrated robust performance only with visually familiar canonical configurations, revealing a significant failure to generalize to even slightly altered, out-of-distribution visual arrangements of the same underlying affordances.](https://arxiv.org/html/2602.05441v1/figures/res_fully_s.jpeg)
New research reveals that even advanced AI-powered robots struggle to reliably perform simple physical tasks when faced with slight variations in their environment.
Mapping Proteins in Space: A New Era for Tissue Biology
![The STProtein training framework leverages a multi-stage approach-initial protein structure prediction, followed by iterative refinement using [latex] \nabla_{\theta} L(\theta, x) [/latex]-to sculpt protein conformations capable of fulfilling designated functional requirements, ultimately demonstrating an adaptive system responding to the inherent entropy of structural possibilities.](https://arxiv.org/html/2602.05811v1/images/structure/framework4.jpg)
Researchers have developed a powerful new method to predict where proteins are located within tissues, even when direct protein measurements are limited.
Closing the Reality Gap in Robotic Manipulation
A new benchmark dataset and simulation environment aims to accelerate the development of more adaptable and reliable mobile robots capable of complex physical tasks.
Unlocking Knowledge: Automatically Defining Scientific Concepts
A new approach uses the power of large language models to systematically extract definitions from the vast body of academic literature.
Logic Circuits Meet Neural Networks: A New Path to Provable Reasoning
Researchers have developed a novel neural network architecture that bridges the gap between the reliability of formal logic and the adaptability of machine learning.
From Goals to Code: A New Approach to Robotic System Design
![Ontological concepts, while seemingly disparate across perspectives, ultimately converge upon a shared foundation of formalized relationships, demonstrable through the consistent application of [latex] \lambda x . \phi(x) [/latex] to define object properties and their interactions within a logically consistent framework.](https://arxiv.org/html/2602.05456v1/ontology_all_perspectives_new.png)
A novel methodology streamlines the development of robotic systems by translating high-level objectives into executable code with formal guarantees.
Can AI Chatbots Show More Empathy Than Doctors?
A new systematic review and meta-analysis suggests that patients often perceive AI-powered chatbots as displaying greater empathy in text-based interactions than their human healthcare providers.