AI Research News

RSS

AI Researchers Sharpen Truth Detection with Fact-Level Hallucination Scanner

Researchers have developed FactSelfCheck, a black-box method that detects AI hallucinations at the fact level using sampling and consistency analysis. This fine-grained approach improves correction accuracy by up to 35% without needing external data or model access.

24 Mar 2025

NdLinear Reinvents Neural Layers To Boost Accuracy And Slash Parameters In AI Models

A new study proposes NdLinear, a multi-dimensional linear layer that preserves data structure and slashes parameter counts across neural networks — from CNNs to Transformers. It outperforms standard layers in accuracy, efficiency, and scalability.

24 Mar 2025

AI-Powered Framework Transforms 3D Metal Printing With Precision and Speed

University of Toronto engineers have developed a machine learning framework, AIDED, to rapidly optimize 3D metal printing settings, reducing trial and error.

24 Mar 2025

Green AI Slashes Emissions but Sacrifices Accuracy in Climate Trade-Off

Training large AI models like GPT-3 generates massive CO₂ emissions—comparable to dozens of transatlantic flights—due to extreme energy demands. Researchers urge a shift to Green AI, which cuts emissions by up to 50% through smarter tech and renewable energy.

23 Mar 2025

New Memristor Mimics Brain Function to Prevent AI From Forgetting

Researchers from Jülich have developed a robust new memristor that mimics brain-like learning and memory. Its novel switching mechanism could overcome the “catastrophic forgetting” problem in neural networks.

21 Mar 2025

AI Outperforms Wall Street: Symbolic Modeling Beats Classic Finance Models in Predicting Market Trends

Texas A&M researchers have created a new AI-driven method called Symbolic Modeling that outperforms traditional asset pricing models by uncovering hidden market dynamics. The model improves prediction accuracy while offering a unified approach to handling diverse financial datasets.

21 Mar 2025

AI Outperforms Humans in Child Interviews by Asking Better Questions

A groundbreaking study found that AI-generated questions using ChatGPT can match or exceed untrained humans in interviewing children about witnessed events. The AI produced more accurate responses per question and followed best practices more consistently.

20 Mar 2025

MIT and NVIDIA Create Lightning-Fast AI That Generates Ultra-Realistic Images

MIT and NVIDIA researchers have developed HART, a hybrid generative AI model that produces high-quality images nine times faster than standard diffusion models. By combining fast autoregressive methods with detailed diffusion refinement, HART runs efficiently on laptops and phones.

20 Mar 2025

New AI Model Fights Malware with Machine-Generated Defense Rules

Researchers are developing a security-focused LLM to combat AI-generated malware, aiming to automate malware detection and defense at scale.

20 Mar 2025

AI Threats Shake National Security: Experts Hesitate or Act Recklessly in Crisis Scenarios

AI-driven threats cause hesitation among security professionals, while those fearing AI replacement respond recklessly, revealing critical challenges in crisis decision-making.

20 Mar 2025

AI Robot Masters Coffee-Making, Paving the Way for Smarter Machines

Researchers at the University of Edinburgh have developed an AI-powered robotic arm that can prepare coffee in a busy kitchen, showcasing human-like adaptability. The breakthrough combines AI, sensitive sensors, and refined motor skills, enabling the robot to navigate unpredictable environments and respond dynamically to real-world interactions.

19 Mar 2025

Wireless Networks Need Human-Like AI to Unlock the Next Tech Revolution

A groundbreaking paper argues that the next revolution in wireless technology hinges on AI with human-like common sense. The researchers propose a paradigm shift, integrating advanced AI with wireless networks to create an "AGI-native" system that can think, plan, and imagine like humans.

19 Mar 2025

AI Predicts Prostate Cancer Treatment Success with 70% Greater Accuracy

AI-driven tumor mapping with Unfold AI helps doctors more accurately assess prostate cancer size, improving treatment success rates for partial gland cryoablation. UCLA researchers found that using AI to measure tumor volume could reduce treatment failures by over 70%.

19 Mar 2025

Researchers Cut AI Energy Use by 90% with Smarter, Leaner Models

Researchers at Saarland University and the German Research Center for Artificial Intelligence have developed techniques to reduce AI energy consumption by up to 90%, making AI models smaller, more efficient, and more accessible to businesses. Their approach, showcased at Hannover Messe, aims to revolutionize sustainable AI.

19 Mar 2025

AI-Powered Ring Revolutionizes ASL Fingerspelling with Real-Time Tracking

Cornell researchers have developed SpellRing, an AI-powered micro-sonar ring that continuously tracks fingerspelling in American Sign Language (ASL) in real time, offering a compact and practical solution for text input and potential ASL translation.

19 Mar 2025

Snap, Scan, and Know: AI Turns Meal Photos into Instant Nutrition Facts

NYU Tandon researchers have developed an AI-powered food recognition system that accurately estimates calories and macronutrients from meal photos, eliminating the need for manual food tracking.

19 Mar 2025

AI Model Analyzes Entire Nights of Sleep to Transform Diagnosis of Sleep Disorders

Researchers at Mount Sinai have developed an AI model, PFTSleep, that analyzes entire nights of sleep with unprecedented accuracy. Trained on over 1 million hours of data, it classifies sleep stages more effectively than traditional methods, paving the way for advanced clinical applications.

18 Mar 2025