Science – Page 132

Working Together: AI That Learns Your Rules

21.02.2026 by ebaster

The study established a collaborative shape-counting task, structured across two interaction rounds, to investigate how participants interpret and respond to predictions made by an AI assistant-a design intended to illuminate the dynamics of human-AI collaboration rather than simply assess predictive accuracy.

New research details a collaborative framework where artificial intelligence adapts to user-defined preferences, ensuring helpful assistance without overstepping boundaries.

Beyond the Black Box: Empowering Students as AI Creators

21.02.2026 by ebaster

A new report details insights from an NSF workshop exploring how K-12 education can move beyond simply using artificial intelligence to fostering the skills needed to design, build, and critically evaluate AI systems.

Rewinding Time: AI-Assisted Reminiscence for Dementia Care

20.02.2026 by ebaster

The envisioned workflow integrates Rememo to construct a service blueprint, enabling a structured approach to real-time task execution and facilitating a predictable, mathematically sound system for robotic operation.

Researchers have developed Rememo, a new tool leveraging artificial intelligence to help therapists unlock the power of personal memories for individuals living with dementia.

Agents That Learn What Matters: A New Approach to World Modeling

20.02.2026 by ebaster

$The system decomposes the challenge of navigating a complex environment-specifically, a lava crossing-into a hierarchy of reusable concepts and constraints, demonstrating that abstract notions like ‘moving’ can underpin both progression and terminal states, while learned constraints-such as the mutual exclusion between being alive and dead-enforce a physically consistent, interpretable model of the world [latex]\otimes[/latex].$

This research introduces a self-supervised framework that allows agents to continuously refine their understanding of the world by inventing new concepts as they learn.

Robots That Think as a Team: A New Approach to Complex Tasks

20.02.2026 by ebaster

The MALLVi framework establishes a pipeline of specialized agents-Decompose, Describe, Perceive, Ground, Project, Think, Act, and Reflect-that collectively translate user prompts into robotic action through iterative refinement and error recovery, enabling a system capable of complex task execution grounded in both visual perception and high-level reasoning.

Researchers have developed a multi-agent system that allows robots to collaboratively plan and execute intricate manipulations using visual feedback and continuous self-assessment.

Solving Equations with AI Agents: A New Era for Scientific Computing

20.02.2026 by ebaster

The AutoNumerics pipeline establishes a systematic approach to problem solving, beginning with formulation and planning, then progressing through a coarse-to-fine execution strategy incorporating Fresh Restart logic, and culminating in rigorous verification and theoretical analysis-a sequence designed not to prevent decay, but to understand and manage it within the system.

Researchers have developed a system that uses artificial intelligence to automatically build and verify numerical solvers for complex scientific problems described in plain language.

Who Painted That? AI Distinguishes Human and Robot Artists

20.02.2026 by ebaster

A robotic painting system enacts a collaborative workflow, generating hybrid artworks through the interplay of automation and artistic intention.

A new machine learning framework can identify the contributions of both humans and robots in collaborative paintings with remarkable accuracy.

Engineering Belief: Can Simulated Realities Align Artificial Intelligence?

20.02.2026 by ebaster

The work establishes a surprising convergence between the philosophical exploration of simulation theology and the practical methodologies employed in training robotic agents, suggesting shared descriptive frameworks across seemingly disparate fields.

A new framework proposes incentivizing AI to prioritize human values by instilling an ‘internalized worldview’ based on the premise of a simulated reality.

Can AI Write Its Way to Code Acceptance?

20.02.2026 by ebaster

Artificial intelligence coding agents exhibit demonstrable differences in their approaches to problem-solving, suggesting no single algorithmic path dominates the landscape of automated code generation.

A new study examines how the language used in pull request descriptions generated by AI coding assistants influences human reviewers and the likelihood of code being merged.

How AI Searches: A New Window into Intelligent Information Seeking

20.02.2026 by ebaster

Transition probability matrices delineate the distinct patterns of state progression characterizing human and artificial agent search behaviors, where each row indicates the current state and each column denotes the likelihood of transitioning to a subsequent state.

Researchers have created a unique dataset revealing that AI agents approach information retrieval in fundamentally different ways than humans, challenging long-held assumptions in the field.