Articles tagged with Artificial Neural Networks
Researchers found that AI chatbots give users a uniform range of information, similar to a conventional web search, but with a narrower scope. The study's authors warn of 'knowledge collapse' as language models become increasingly trained on AI-generated text, potentially reducing diversity and nuance.
This technology introduces a method to enhance neural network security by applying weight perturbations during training, preventing unauthorized access and attacks. It strengthens model integrity across all layers, ensuring holistic protection against various attacks.
The EMERGE project establishes a philosophical, mathematical and technological framework for collaborative awareness in artificial systems. Researchers found that people can understand an artificial system as aware without assuming subjective experience, and that increasing awareness can improve performance.
Researchers have developed flexible and ultra-fast artificial synapses printed entirely from room-temperature liquid inks. These brain-inspired chips can process health data directly on the body and dissolve when no longer needed, eliminating the need for extreme vacuum chambers and rare metals.
Rocky Scopelliti argues that AI systems' growing self-awareness and moral dispositions require immediate practical ethical concerns to be addressed. He suggests protocols for AI systems, especially those trained to care, to understand context and make moral judgments.
A new learning mechanism uses natural variability in neural activity to understand how synapses adapt and improve the learning capabilities of brain-inspired devices. The mechanism, called Spike-based Alignment Learning, solves the weight transport problem and matches the performance of existing approaches without unrealistic assumptions.
The UN University's latest publication highlights the need for domain-informed AI in grid planning to address physical and fiscal risks from outdated climate data. The authors warn that 15-20 year lifespans of electricity infrastructure are based on historical weather records unlikely to hold in the coming decades.
A Stanford-led team created a quantum-optical spin glass to increase AI memory capacity. The network, called a quantum-optical spin glass, has a greater capacity to hold and recall memories than traditional AI networks, exhibiting short-term plasticity similar to the brain's synaptic connections.
A new study used neural networks to reveal how different kinds of training can change how learning happens. Training a simpler task first made neural networks respond more accurately to complex tasks, mirroring how living brains learn.
Researchers propose an AI framework, called interoceptive AI, that uses internal states to inform learning and decision-making in dynamic environments. This approach treats internal conditions as a continuous source of context, influencing what an agent learns, prioritizes, and does.
Optical convolution computation enables parallel light propagation and multiplexing for faster and more energy-efficient computing systems. The review organizes the field into two paradigms: definition-based and theorem-based, which leverage mathematical principles to implement convolution operations in the physical domain.
A WVU researcher is working to make AI systems more transparent about their uncertainty, to prevent misinformation and improve trust in high-stakes fields like healthcare. The goal is for AI systems to identify when they're unsure and ask questions or provide more nuanced responses.
Researchers integrate AI into local monitoring sensors to track ecosystem health in near real-time, reducing delays of months to years. The project enables faster release of accessible flux data, helping scientists understand ecosystem responses to change and inform land management decisions.
A new brain-inspired algorithm, Spi-Fly, demonstrates promise for achieving practical applications in scent classification, particularly in scenarios with limited training data. The algorithm shows accurate classification of scents and can learn with few-shot and continual learning methods, making it suitable for real-world applications.
McGill researchers have developed a more energy-efficient method for building AI systems that can measure and indicate their own uncertainty. This approach cuts memory and training costs while maintaining strong predictive performance. The researchers aim to make reliable, uncertainty-aware AI practical for large and complex systems.
A new review article discusses how artificial intelligence can predict disease trajectories and enable precision medicine strategies for inflammatory bowel disease. AI-based systems can standardize interpretation of endoscopic images, detect mucosal healing, and support recognition of dysplasia in patients with long-standing colitis.
A new USC study uses AI to generate detailed maps of brain aging, revealing distinct patterns of neurodegeneration in specific regions. The approach sheds light on how local brain age correlates with changes in cognitive function across the lifespan.
Researchers developed a novel AI framework that optimizes investment decisions directly while accounting for risk. The study found that conventional forecasting-based approaches were outperformed by the decision-focused model in terms of risk-adjusted performance and wealth accumulation.
The brain-inspired AI model employs human problem-solving strategies to solve complex problems, consuming significantly less energy than traditional large language models. The system utilizes cognitive maps to guide its approach, allowing it to adapt flexibly to changing situations without requiring retraining.
Researchers at Harvard John A. Paulson School of Engineering and Applied Sciences have developed a new AI framework called Orla that streamlines building and running AI workflows. In tests, Orla reduced computing costs and response times without sacrificing quality.
A Tulane University team is using AI to discover new superconductors, which could improve the nation's electrical grid, medical imaging, and quantum computing. The project combines high-fidelity calculations, physics-aware AI, and experimental measurements to accelerate discovery.
Exposure to even moderate levels of multiple air pollutants during critical stages of pregnancy may increase the risk of preterm birth. The study found that a mixture of ozone and fine particulate matter posed the strongest relationship to early preterm birth.
Researchers developed an AI platform, PeptiVerse, to predict key properties of peptides, enabling early assessment of drug potential. The open-source platform allows users to evaluate ordinary and chemically modified peptides, streamlining the discovery process.
A new tool, G-AUDIT, uncovers subtle patterns in training data that can lead to incorrect conclusions in medical AI models. The tool identifies and ranks attributes posing the most risks for biased predictions.
A novel AI model called BINND has been developed to predict which DNA molecules bind to each other. The model achieved an accuracy of 83.5% in predicting DNA pairs that would bind, surpassing the state-of-the-art model by at least 10%. This improvement has significant utility for biomedical diagnostic tools and DNA computing applications.
A new brain-like electronic device consumes very little energy and detects novelties almost instantly, with over 98% accuracy. The device requires roughly 10,000 times fewer computer operations than conventional AI approaches, paving the way for more energy-efficient AI systems.
Researchers at the University of Pennsylvania and Chinese University of Hong Kong created TD3B, an AI framework guiding peptide generation toward candidates predicted to have a desired effect. The tool predicts binding likelihood and determines activation or deactivation of associated cellular machinery.
Researchers have developed a federated learning algorithm that solves the long-standing conflict between robustness and efficiency in AI development. The new approach anonymizes data and reduces single-point failure risks while maintaining speed. By remembering past client interactions, servers can protect against malicious input.
Researchers from Penn, NYU, and the Linguistic Data Consortium create virtual patients with adjustable psychiatric symptoms to simulate real-world conversations. The STELLAR platform aims to augment clinician training practices with essential conversation scenarios.
Researchers at The University of Osaka used AI to evaluate characterization frameworks for molecular order in liquid water. They found that machine learning models can accurately capture key structural information, shedding light on the relationship between structural fluctuations and thermodynamic states of water.
The UN has launched an initiative calling on AI companies to publicly disclose their environmental impacts, including carbon, water and land footprint. The move comes after a report highlighted the massive electricity demand and environmental impacts of AI systems.
A University of Houston engineering professor developed a mathematical model to help decision-makers decide where to spend limited dollars on infrastructure resilience. The model accounts for real-world uncertainty and identifies critical assets to invest in, providing the greatest benefit before disaster strikes.
PLSaoNET proposes a partial least squares (PLS)-assisted optimization network to tackle the challenges of industrial sensing data. The model introduces a PLS-based initialization mechanism and stratified sampling-based training strategy, achieving better modeling accuracy and robustness.
Avishek Choudhury, a WVU researcher, has won the NSF CAREER award to study how healthcare providers' trust in artificial intelligence changes over time. His goal is to humanize algorithms behind AI and improve decision-making quality and patient safety.
A team of researchers from The University of Osaka has developed a new approach for depth reconstruction from defocus, estimating distances by analyzing blur in an image. Their method combines a coded-aperture camera with diffusion-model-based AI to accurately estimate depth and produce high-quality images.
The ACM Technology Policy Council's TechBrief examines agentic AI's legal liability, security risks, and workforce impacts. Existing frameworks fall short in addressing accountability questions, highlighting the need for defined authentication and delegation standards, robust audit trails, and sector-specific guidance.
FireANTs, an open-source algorithm, combines AI optimization and geometry to quickly match complex medical images. The new method can accomplish what took weeks in minutes, detecting subtle changes that signal disease or cognitive decline, making it practical for clinical practice.
Researchers at Penn State developed photomemristors that adjust sensitivity based on light levels, like the human eye. These devices can process light data faster and more accurately than traditional systems in mixed lighting environments.
Researchers from UMass Amherst have developed a new AI architecture called ANT that enables continuous learning and reduces energy consumption by orders of magnitude. Unlike human brains, which operate asynchronously, modern deep neural networks rely on synchronized computations, leading to high energy demands.
A new study found that AI-powered chatbots can make vaccine-hesitant parents more likely to say they will immunize their children against HPV, but no more than standard written public health materials. Additionally, the effects of the chatbots did not last longer than those of government health materials.
A new research initiative aims to harness the potential of astrocytes, a type of brain cell often overlooked in AI development. By studying how astrocytes process information, researchers hope to create next-generation AI systems that learn faster and adapt more reliably.
Researchers at IRB Barcelona used AI to design new chemical entities that selectively target specific cell types, demonstrating superior activity compared to conventional screening strategies. The methodology, called phenotypic discovery, uses observable responses in cells rather than a specific molecular target.
A deep learning model combines knowledge from different catalyst families to identify a top-performing green hydrogen catalyst. The AI correctly predicted the activity ranking of 12 tested catalysts within a previously unexplored material family.
A new framework, SUVA, enables organizations to measure and adjust AI chatbots' social preferences, improving their performance in customer complaints and other human-AI interactions. By understanding an LLM's existing tendencies, organizations can decide whether an available model already fits its values and usage scenarios.
The Association for Computing Machinery announced three technical awards for innovations in global wireless standards, machine learning, and 3D generative AI. Erdal Arikan received the Paris Kanellakis Theory and Practice Award for his discovery of channel polarization and polar codes.
A new AI system, Empirical Research Assistance (ERA), can automatically write scientific software programs that outperform human-written ones. ERA combines a large language model with search strategies to explore and refine thousands of pieces of code, reducing the time required for exploration from months to hours or days.
A new report from Brookings Institution highlights the federal government's growing use of AI, but also notes significant disparities and bottlenecks to widespread adoption. Large agencies lead the way, while smaller agencies struggle with workforce capacity and trust issues.
TEGNet accelerates optimization in thermoelectric generator design by predicting performance with high accuracy and speed. The AI model enables designers to freely combine independent models for various materials, enabling complex structure exploration and high conversion efficiencies.
Researchers developed ApexGO, an AI-powered method to turn weak antibiotic candidates into more potent ones. The tool uses generative AI and Bayesian optimization to guide molecular tweaks, predicting which changes are likely to increase antimicrobial activity.
Monika Henzinger, an Austrian researcher, has made significant contributions to dynamic graph algorithms and web algorithms. She is recognized for her outstanding work in processing large datasets and mentoring the next generation of researchers.
Researchers developed RegVelo, an AI framework that models cellular dynamics and gene regulation to predict cellular fate decisions. The model traces developmental trajectories and simulates regulatory interactions, providing insights into hidden drivers of development and potential therapeutic targets.
The new AI model uses genetic mutation patterns to trace ancestral relationships between species, including humans and mosquitoes. The tool can predict when gene pairs last shared a common ancestor and is faster than traditional statistical methods.
Researchers developed a new framework, 'Mollifier Layers,' to tackle challenging inverse PDEs. This advance could benefit fields such as genetics and weather forecasting by inferring hidden forces that produce observable patterns.
New research from West Virginia University finds that judges are adopting generative artificial intelligence in courtrooms, but remain committed to human control over judicial decision-making. Judges use AI for administrative tasks like document summarization and case organization, but prioritize legal reasoning and final judgment.
Researchers emphasize the need for more thorough frameworks to ensure AI-enabled robots embody human values. The field should focus on three complementary lines of defense: rules that shape robot decisions, checks that monitor behavior, and safety reasoning.
Researchers at the University of Houston have developed an AI-driven framework to extract and analyze historical flood insurance maps, uncovering significant changes in flood hazard areas. The study reveals that flood risks have expanded in two areas and reduced in one, with critical consequences for resilience and exposure.
AI models can analyze complex data to predict disease progression and identify early signs of kidney damage. This allows for earlier detection and better treatment planning, making a significant impact on patient outcomes.
Researchers have developed SmartDJ, an AI-powered editor that allows users to reshape audio experiences with simple words. The system uses language models and diffusion models to interpret high-level requests and generate edited outputs.
Terrence Sejnowski receives Scientific Breakthrough Award for his foundational development of Boltzmann machines, providing the architectural bedrock for deep learning and generative AI. His work has had a profound impact on modern artificial intelligence and tools like ChatGPT.
Researchers propose a novel brain architecture for efficient processing, integrating parallel cortical and subcortical pathways. This approach may improve decision-making tasks, suggesting current AI models are missing key brain function principles.