Light Speed Microbots: Harnessing Refraction for Particle Propulsion
New research demonstrates how carefully shaped microparticles can be propelled by light refraction, paving the way for remotely controlled micro-robotic systems.
Science
From Physics Papers to Executable Code: An AI-Powered Leap for HEP Analysis
Researchers are exploring how large language models can automatically translate the methods described in high-energy physics publications into functional code.
Smart Homes Need to Adapt: Testing Occupancy Detection Across Apartments
A new study rigorously evaluates how well machine learning models trained in one home can accurately detect occupancy in others, revealing crucial insights for truly smart building technology.
Smart Deployment: Automating AI at the Edge
A new framework leverages intelligent agents to simplify and accelerate the deployment of AI models, particularly for resource-constrained devices.
Sharing Insights, Not Data: A New Approach to Personalized Recommendations
![The system architecture addresses behavior prediction through a two-stage process: initially, item semantics are aggregated-with privacy safeguards-and converted into ID embeddings via Federated Knowledge Distillation [latex]FKD[/latex]; subsequently, this pre-learned knowledge is projected into prompts for fine-tuning a Large Language Model [latex]LLM[/latex], enabling predictive capabilities as the system gracefully ages.](https://arxiv.org/html/2604.14833v1/x1.png)
Researchers are exploring methods to build effective recommendation systems without compromising user privacy, leveraging the power of distributed learning and advanced language models.
Beyond Real: AI Synthesizes MRI Data to Enhance Brain Imaging
![A conditional variational autoencoder compresses complex-valued MRI patches [latex]2\times 96\times 96[/latex] into latent representations [latex]2\times 48\times 48[/latex], subsequently employed to train a flow matching model in two stages-first learning sequence-specific brain anatomy, then refining its understanding through conditioning on labeled normal and abnormal data.](https://arxiv.org/html/2604.14800v1/pipeline.png)
A new generative modeling framework creates realistic brain MRI scans, unlocking potential for improved diagnostics and reduced reliance on invasive procedures.
Rewiring the Brain: Continual Learning for Faster, More Accurate Diagnosis
A new framework tackles the challenges of applying machine learning to fMRI data, enabling models to learn from diverse datasets without forgetting previous knowledge.
Robots That Remember: Smarter Object Search in Cluttered Spaces
A new framework combines probabilistic planning with neural networks to help robots efficiently locate objects even as environments become more complex.
Guiding Language Models with Beliefs and Policies
![Credo enables the declaration of logical pipelines, beliefs, and policies via its APIs, after which an execution engine dynamically rewrites and enacts these pipelines, adapting to runtime beliefs and matched policies to ensure consistent and informed action-a process grounded in the principle that behavior stems from logically derived, belief-driven execution [latex] B \rightarrow P \rightarrow A [/latex].](https://arxiv.org/html/2604.14401v1/x2.png)
A new system decouples pipeline logic from execution, allowing language model workflows to dynamically adapt based on semantic understanding.
Light-Driven Microbots Take Shape
Researchers have designed and 3D-printed microparticles that move autonomously when illuminated, opening doors to new controllable active matter systems.