Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
HoprLabs, a new open-source Python CLI and research toolkit, launched to simulate AI training math, estimating model size, memory usage, training time, and token budget before costly model training. The MIT-licensed tool…
An AI-powered behavioral classifier detected a customer's hesitation about battery life during a laptop purchase and restructured the product page in real time, converting a £1,099 sale in 44 seconds. The system, built o…
Scaling laws in deep learning describe a predictable power-law relationship between training loss and model size, dataset size, and compute, enabling extrapolation of resource requirements for larger models. Early theore…
A developer spent three iterations building an auto-fix pipeline that still only works reliably on trivial tickets. The pipeline, which pulls tickets from Azure DevOps, runs them through a local model, and pushes fixes v…
A developer built Rahnuma.io, an AI-powered DevOps platform that combines project management with developer workflow understanding, addressing the gap between heavy tools like Jira and light tools like Trello. The platfo…
A developer built CORE, a cognitive system using Cl(4,1) Conformal Geometric Algebra to make incoherence algebraically impossible. Unlike large language models that monitor and correct for coherence, CORE enforces a vers…
Seven AI-powered trading tools are democratizing sophisticated market analysis and automation, enabling retail traders to spot chart patterns, build strategies, and automate trades without deep coding skills. Platforms l…
Neo4j Chief Scientist proposes that AI agents require three types of structured memory—long-term, short-term, and reasoning memory—organized as context graphs to ensure reliability and accuracy in autonomous systems. The…
Researchers used AI and high-resolution CT scans to read a sealed scroll from Herculaneum, destroyed by Mount Vesuvius nearly 2,000 years ago. The scroll, PHerc. 1667, contains a Stoic philosophy treatise on ethics, mark…
A developer compiled FFmpeg from source with NVENC/NVDEC hardware acceleration on an NVIDIA Jetson AGX Orin 64GB running Ubuntu 22.04 LTS and JetPack 6.2.2 (CUDA 12.6). The custom build offloads video encoding and decodi…
Neurometric AI launched an automated token engineering platform and raised $4 million in a pre-seed round from investors including Betaworks, ex/ante, and Jason Calacanis. The platform optimizes AI model calls, reducing …
Topos, an open-source tool from Krv-Labs, evaluates code quality by mapping programs to structural graphs and calculating metrics for simplicity, composability, and security. It aims to address the bottleneck of human co…
An automotive engineer faces 26-hour design iterations for CFD simulations of a side mirror, motivating a new generation of AI tools like PhysIQ that aim to accelerate physics simulation. This primer explains the basics …
Robust.AI selected Aptiv PLC's intelligent perception systems, including the Aptiv PULSE sensor, for its Gen 3 Carter collaborative mobile robot. The sensor fuses radar and vision using machine learning to improve reliab…
Revolut has developed PRAGMA, a transformer-based banking foundation model trained on 40 billion events from 25 million users, achieving a 64.7% improvement in fraud recall and up to 130% uplift in credit scoring metrics…
A Walmart software engineer published InfiniteDiffusion, an independent paper accepted to SIGGRAPH 2026, introducing a new approach to infinite generation with diffusion models and the world's first learned procedural te…
The softmax function, essential to modern AI systems like GPT, originated from Ludwig Boltzmann's 19th-century work on thermodynamics, where he derived the exponential form to maximize entropy in gas molecules. The same …
Researchers introduced Autodata, a method enabling AI agents to act as data scientists that build high-quality synthetic training and evaluation data. The approach, including a practical implementation called Agentic Sel…
Capital One appointed Prem Natarajan, former head of Alexa AI at Amazon, as its Chief Scientist to lead AI research and deployment in financial services. The bank aims to solve domain-specific challenges like real-time f…
Red Alice AI released the first official benchmark of its Version 2 architecture, reporting a 200x performance gain in the RedTensor engine. The upgrade introduces a PyTorch-backed TorchTensor backend with unified flat i…