Skip to content

BBG NEWS

  • Science
  • Who is Denis Avetisyan?

Science

AI Uncovers Hidden Equations of Nature

12.02.2026 by ebaster

The PiT-PO framework refines large language models into adaptive equation generators through a reinforcement learning process driven by dual-constraint evaluation-balancing physical feasibility with theoretical consistency-to produce parsimonious and physically plausible results at a token level, effectively moving beyond static proposition towards dynamic refinement.

Researchers are leveraging the power of artificial intelligence to automatically discover fundamental equations governing complex physical phenomena.

Categories Science

Seeing the Crowd: Smarter Paths for Robots

12.02.2026 by ebaster

Visual Language Models demonstrate capacity for crowd recognition, enabling quantitative analysis of population density and distribution.

New research leverages advanced AI to help robots understand and navigate complex environments with people, going beyond simple obstacle avoidance.

Categories Science

Beyond the Algorithm: How AI Can Sharpen, Not Replace, Human Judgment

12.02.2026 by ebaster

A system assesses decision-making by adopting the perspective of the decision-maker, evaluating confidence levels, and then prompting critical self-reflection through counterfactual analysis-adding or removing features from the initial reasoning-to facilitate AI-supported revisions and data-driven validation of conclusions.

New research explores how AI-powered tools can foster deeper critical thinking in decision-making, moving beyond simple reliance on automated outputs.

Categories Science

Robots Learn to Imagine: Scaling Manipulation with World Models

12.02.2026 by ebaster

RISE distinguishes itself from existing reinforcement learning approaches-which typically depend on data gathered from prior, real-world interactions-by facilitating on-policy learning through the construction of a world model that functions as a dynamic, interactive environment for policy optimization.

A new framework allows robots to refine their skills in a simulated environment, dramatically improving performance on complex physical tasks.

Categories Science

Looking Up: AI and the Future of Astrophysics

12.02.2026 by ebaster

As artificial intelligence rapidly evolves, the field of astrophysics must grapple with its potential benefits and risks to ensure continued scientific progress.

Categories Science

Beyond Chatbots: The Rise of AI Agents

12.02.2026 by ebaster

The research demonstrates a progression from conventional, passive models toward the development of goal-directed systems capable of intentional behavior, signifying a shift in robotic control paradigms.

A new wave of artificial intelligence systems is moving beyond simple question-and-answer interactions to proactively pursue complex goals.

Categories Science

Robots Learn by Watching: Predicting Motion for Flexible Imitation

12.02.2026 by ebaster

Flow accumulation, as measured by both average displacement error (ADE) and final displacement error (FDE), decreases on a logarithmic scale across five independent experimental seeds.

New research demonstrates a method for robots to learn complex tasks from limited human demonstrations by focusing on predicting 3D scene flow and leveraging cropped point cloud data.

Categories Science

Ask and You Shall Discover: AI Agents Chart a New Course for Materials Science

12.02.2026 by ebaster

The Materials Knowledge Navigation Agent presents a user interface enabling researchers to pose open-ended scientific questions, customize data retrieval and filtering criteria, and execute each step of an automated analysis pipeline with interactive control.

A new autonomous agent leverages the power of natural language to navigate vast scientific literature and accelerate the discovery of materials with targeted properties.

Categories Science

Leap of Faith: Robots Learn to Jump Together

12.02.2026 by ebaster

The Co-jump framework cultivates collaborative locomotion through a multi-agent proximal policy optimization (MAPPO) architecture-employing independent policy and value networks for each robot-and a four-stage curriculum learning progression-beginning with gravity adaptation, progressing to target acquisition, then initialization refinement, and finally, delay compensation-demonstrating its practical implementation on quadrupedal robots.

Researchers have developed a reinforcement learning system that allows two quadruped robots to collaboratively perform complex jumping maneuvers, pushing the limits of robotic teamwork.

Categories Science

Beyond the Bots: Orchestrating AI for Real-World Impact

12.02.2026 by ebaster

Specialized artificial intelligence workflows, each dedicated to a distinct business function, are integrated under human supervision via multi-channel presentation interfaces, enabling coordinated orchestration of complex processes and demonstrating a synergistic human-AI collaboration.

Successfully integrating intelligent agents into the workplace demands a strategic shift towards workflow design and human oversight, rather than simply deploying the latest AI technology.

Categories Science
Older posts
Newer posts
← Previous Page1 … Page48 Page49 Page50 … Page104 Next →
© 2026 BBG NEWS • Built with GeneratePress