Science – Page 89

Robots in Tight Spaces: A New Approach to Motion Planning

01.03.2026 by ebaster

A multi-link robot navigates a constrained, U-shaped passage by dynamically reshaping its structure through a hierarchical trajectory planning framework, demonstrating an approach to robotic locomotion that prioritizes adaptable morphology for complex environments.

Researchers have developed a hierarchical planning framework that allows complex, floating-base robots to navigate challenging, confined environments with greater efficiency and reliability.

Ask the Expert: Smarter AI Through Human Collaboration

01.03.2026 by ebaster

The system addresses failure in complex tasks by bridging two distinct knowledge gaps-a lack of foundational factual understanding (such as attempting a task without the necessary tools) and a deficiency in higher-level strategic reasoning-through the dynamic integration of human expertise when and where it is needed to ensure successful completion.

A new framework empowers AI agents to seek guidance from humans, overcoming knowledge gaps and achieving improved performance in complex tasks.

Democratizing Dangerous Biology: How AI Lowers the Barrier to Biosecurity Risks

01.03.2026 by ebaster

Leveraging large language model assistance significantly improves the performance of novice users on complex tasks, frequently achieving accuracy levels comparable to, and sometimes exceeding, those of both the language models themselves and seasoned human experts.

New research reveals that artificial intelligence is making complex biological tasks accessible to a wider range of actors, raising concerns about the potential for misuse.

Smarter AI: Cutting Energy Costs with Adaptive Language Models

01.03.2026 by ebaster

A new approach to large language model inference dynamically selects model size based on query complexity, offering significant energy savings without sacrificing performance.

Beyond Predictions: How AliExpress is Using AI to Understand What Shoppers Want

01.03.2026 by ebaster

SIGMA establishes a holistic framework wherein system behavior arises from underlying structure, prioritizing clarity and simplicity as foundational elements of effective design and control.

A new generative AI system is moving beyond simply predicting purchases to actively interpreting user intent and tailoring recommendations accordingly.

Boosting AI Detector Reliability with Smart Data Fusion

01.03.2026 by ebaster

$The system contrasts conventional supervised learning-with distinct training and inference phases-against a proposed inference method utilizing resampling to generate multiple samples from a single observation, notably excluding [latex]\text{Char}(f)[/latex] from model inputs when the characteristic function is unknown to enhance robustness and explore solution space.$

A new technique combines the outputs of AI-based wireless detectors to significantly improve accuracy and reduce prediction uncertainty.

Storytelling Bots and Resilient Minds

01.03.2026 by ebaster

The TaleBot interface facilitates exploration of three distinct interaction modes, enabling a system to navigate varied conversational landscapes and dissect the nuances of dialogue.

Researchers explore how AI-powered interactive narratives can help elementary school children develop crucial coping skills.

Building Trust in the Age of AI

01.03.2026 by ebaster

New research reveals that a culture of psychological safety is crucial for getting employees to initially embrace artificial intelligence technologies.

Keeping AI on Track: Contracts for Reliable Autonomous Agents

28.02.2026 by ebaster

AgentAssert’s contract enforcement introduces a negligible runtime overhead-scaling linearly with constraint count [latex]k[/latex] and remaining under 15 ms for [latex]k=50[/latex] and 25 ms for [latex]k=100[/latex]-a performance margin substantial enough to remain imperceptible relative to the 1,000-3,000 ms latency inherent in large language model inference.

A new framework offers quantifiable guarantees for the safe and predictable operation of AI agents, mitigating risks associated with behavioral drift and ensuring responsible AI governance.

Mapping Life’s Code: How AI Uncovers Hidden Structure in Cells

28.02.2026 by ebaster

As a transformer model deepens, its internal geometry progressively concentrates gene representations along a secretory/localization axis-evidenced by a rise in explained variance from 19% to 77%-while simultaneously embedding co-regulated genes in close proximity, as demonstrated by consistently significant co-localization of TRRUST regulatory pairs across all layers.

New research reveals that artificial intelligence models are organizing gene data into a biologically meaningful framework, offering unprecedented insight into cellular organization and regulation.