Mapping Organic Crystal Landscapes with Generative AI
Researchers have developed a new AI framework that accurately predicts the arrangement of molecules in organic crystals, a critical step for designing advanced materials.
Science
Beyond Imitation: Teaching Robots to Learn and Adapt
A new reinforcement learning framework enables humanoid robots to acquire versatile motor skills by combining reference guidance with goal-conditioned learning.
Robots That Reason: Bringing Common Sense to 3D Manipulation
A new framework empowers robots to understand and perform complex physical tasks in three dimensions by leveraging the power of large language models and multi-agent reasoning.
Seeing, Thinking, and Acting: A New Path for Robotic Reasoning
![Halofirst anticipates task completion not through direct instruction, but by cultivating a self-attentive ecosystem of specialized experts - a multimodal understanding module, a visual generator, and an action predictor - working in concert to infer visual subgoals and execute actions conditioned on emergent, contextual reasoning [latex]EM-CoT[/latex] within a Mixture-of-Transformers architecture.](https://arxiv.org/html/2602.21157v1/x1.png)
Researchers have developed a unified model that allows robots to better understand complex instructions and perform intricate tasks by combining visual perception, language understanding, and action planning.
Beyond Self: Modeling Empathy in Artificial Agents
![The landscape of mutual cooperation shifts dramatically with even subtle variations in dyadic empathy, as simulations reveal that the fraction of collaboratively successful rounds-quantified as [latex] (C,C) [/latex]-is acutely sensitive to the empathy parameters [latex] \lambda_{i} [/latex] and [latex] \lambda_{j} [/latex] of interacting agents.](https://arxiv.org/html/2602.20936v1/images/Fig1_agent_empathy.png)
New research demonstrates how incorporating models of others’ preferences into planning algorithms can foster cooperation and unlock more nuanced interactions between artificial agents.
Bridging the Reality Gap: Sim-to-Real Robot Learning Takes a Leap Forward
New research demonstrates that standard reinforcement learning techniques, when carefully tuned, can reliably transfer policies learned in simulation to physical robots for stable and efficient online learning.
Beyond Automation: Building Truly Skilled AI Agents
A new systematic review explores the emerging concept of ‘agentic skills’ and how reusable procedural knowledge is key to unlocking the next generation of powerful and reliable AI agents.
Swarm Intelligence: How Robots Spread Information by Chance Encounters
![The study demonstrates that the interaction interval within a robotic swarm is governed by a relationship between the number of robots ([latex] NN [/latex]), communication range ([latex] CC [/latex]), and the dimensions of the operational environment ([latex] LL [/latex]), all considered in relation to the robots’ velocity ([latex] vv [/latex]), as evidenced by reported means and inter-quartile ranges.](https://arxiv.org/html/2602.21148v1/x4.png)
New research details how information propagates through robot swarms relying solely on direct, opportunistic communication between individuals.
Decoding Model Uncertainty: New Laws for Biological Insights
![A hierarchical framework for analyzing parameter identifiability integrates eigenvalue decomposition and the Schur complement to categorize parameters across scales, revealing that predictive uncertainty stems from non-identifiable subspaces-specifically, contributions from zero-order non-identifiable parameters [latex]\boldsymbol{U\_{k-r\_{0}}^{\to p}\theta}[/latex] and first-order non-identifiable parameters [latex]\boldsymbol{U\_{k-r\_{0}-r\_{1}}^{\to p}\theta}[/latex]-and quantifying this uncertainty through a metric [latex]\mathcal{K}\_{i}[/latex] that defines practical identifiability.](https://arxiv.org/html/2602.20495v1/x1.png)
A new computational framework reveals how accurately we can estimate parameters in complex biological models, paving the way for more reliable data-driven discoveries.
Teaching Robots to Learn from Interaction
A new reinforcement learning framework guides vision-language models to master complex, long-duration manipulation tasks through interactive feedback.