Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
Dyna Robotics introduced DYNA-2, a robot foundation model trained on more than 1 million hours of first-person human video, on August 10, 2026. The company reported that success rates on high-precision manufacturing task…
Tensordyne (formerly Recogni), an AI system solution company, is hiring a Junior Software Engineer for ML Runtime in Rust, based in Munich, Germany, with an on-site work type. The role involves building and optimizing co…
Google's Gemini API File Search and Microsoft 365 Copilot are expanding toward multimodal RAG, enabling AI assistants to retrieve and cite visual evidence such as charts, screenshots, and diagrams, according to a develop…
A developer with 25 years in enterprise infrastructure argues that graph engineering is essentially a new name for entity-relationship thinking, a core discipline of enterprise architecture. The developer claims that con…
A developer built Visual Forensics Radar, an 'Ensemble of Experts' system that combines Error Level Analysis, Zero-Shot CLIP classification, and a Vision-Language Model (Qwen2-VL) to detect hybrid deepfakes. The system a…
OpenRouter launched a major update to its Auto router (openrouter/auto) that uses market spend data from over 55 trillion weekly tokens to select models, claiming it outperforms the previous router across tasks and cost …
Gusto is hiring a Software Engineer for its ML Platform team in Denver, New York, or San Francisco, offering a salary of $160k–200k/yr. The role involves building and scaling MLOps infrastructure, with a focus on AI-assi…
Gilead Sciences Inc. director of global product security Thomas Luu said his team uses graph neural networks with Neo4j to surface hidden fraud networks in pharmaceutical anti-counterfeiting, detecting schemes that rules…
OpenParser, an open-source API from Eigenpal, integrates nine OCR models from PaddleOCR, Mistral, Azure, Google, and AWS Textract into a single interface, allowing developers to swap models without rewriting code. The pr…
A developer traces the evolution of two-stage object detectors from R-CNN's 47-second inference time to Faster R-CNN's near-real-time 5 FPS, highlighting key innovations like RoI Pooling and the Region Proposal Network. …
A new analysis of vision-language models (VLMs) shows they fail basic physics reasoning, with the average model scoring around 40% on the PhysBench benchmark (ICLR 2025) and the best model, GPT-4o, scoring 49.49%, compar…
An engineer detailed how AI coding agents like Claude Code struggle with data science pipelines due to invisible state and memory limitations, and introduced oryxflow, a lightweight workflow library that uses task depend…
NVIDIA released Magpie Multilingual TTS, an open-weights text-to-speech model with 364 million parameters supporting 12 languages, including new additions Modern Standard Arabic, Korean, and Brazilian Portuguese, and ava…
Mercor, a profitable Series C AI data company valued at $10 billion, is hiring a Member of Technical Staff for Deeptune, its RL environments lab, with a salary range of $220k–600k/yr, based on-site in New York City. The …
Malachyte, an AI-powered ecommerce recommendation platform founded by Sidd, uses Google Cloud's Bigtable and Managed Service for Apache Kafka to help retailers double or even triple their sales by solving the cold-start …
A developer built StatMate (getstatmate.com) and argues that economists are well-suited for data science because the core statistical methods, such as linear and logistic regression, are the same as those in econometrics…
HypoLab, an open-source agentic pipeline built by an unnamed developer, lets an LLM propose data hypotheses but requires classical statistical tests to verify them before presenting results. In tests on the Gapminder dat…
An engineer working on AI model evaluation and fine-tuning reports that accuracy is the most common and subtle failure point in model outputs, often manifesting as confident but factually wrong errors. The engineer highl…
A new explainer by an unnamed researcher details how to measure forecast calibration using empirical coverage and Probability Integral Transform (PIT) histograms, drawing on a 2007 paper by Gneiting, Balabdaoui, and Raft…
The third International Olympiad in Artificial Intelligence (IOAI) concluded in Astana, Kazakhstan, on 8 August 2026, with 108 participating countries and territories, 466 contestants, 126 team leaders, and 100 observers…